00:00 - Lena Robinson (Host)
Hi Everyone, Welcome to Creatives WithAI. I'm your host, Lena Robinson, and today we have a really good friend of mine, somebody I've known, for…I was working out…over a decade, I think, well over a decade. Benjamin Southworth. Welcome, Ben.
00:17 - Benjamin Southworth (Guest)
Hi, thanks for having me, Lena. I really appreciate it.
00:19 - Lena Robinson (Host)
I’m excited to be talking to you because you've been somebody that's been in the tech scene for a really long time, not just from a technology specific thing, but also entrepreneurship in the VC world and founding lots of different businesses. Some of you may know Ben as one of the original ‘Three Beards’, when they used to have the ‘Silicon Drinkabouts’ and the ‘Don't Pitch Me Bros’ and all the different events. You know, community has been something that you've been building around technology for a long, long time, but you also are a very creative thinker, and you know you, you are an artist, you are a musician, you've worked in art galleries. You've worked in technology from the perspective of being the Deputy CEO of Tech City, which I think is now called Tech Nation. So, today's conversation, I think, will be a really interesting one, because you have been at the forefront of what the future of technology looks like for London, and the world really, with regards to humans and community, for quite a long time. And I think, yeah, the first question I'd really like to talk to you about is what you think creativity is? What's your view on creativity?
01:48 - Benjamin Southworth (Guest)
So, I have. I have quite sort of strong opinions about this, and so one is a quote I believe it’s Picasso, which says “the artist is the child that survived”, and I always really enjoy that idea that ultimately any artistic endeavour is actually just play, and it's the ability to just hold on to that sort of childlike wonder of the world, of people, of everything, and to sort of try to not become cynical. And that, as I think all of us probably know, by the time you reach your 40s and you've been through election hopes and sadnesses that you know, remaining optimistic can be difficult in the face of adversity. And then also, I was so lucky that I grew up with two parents who are artistic. You know, who are artists first and teachers second. So, my father is a Professor of Education. He used to be a teacher. My mum is a teacher, mainly dealing with, sort of young children. But my mother is incredibly crafty, huge into her fabrics, her gardening, and my father is an exceptional watercolourist. Mainly of architecture, especially around sort of places. We travelled in France and things as a kid on camping holidays.
03:24
So, I've been, I've always been encouraged, and I've always been rewarded for doing art, I suppose right, as you know. So, it was encouraged, it was and it was. You know it's inspiring to have people around you who are good. I think you know that phrase you can't be what you can't see is really important. And so, I was lucky that my parents read that they took me to museums, that they put me into arts clubs and theatre schools and music lessons and all of it. They really threw me in the deep end of things they thought I would be good at. I mean, they also encouraged the sports. But, you know, I was so lucky to grow up, you know again, middle class, as the son of a Professor in Cambridge. You know, very lucky really and very, very privileged. So, you know, half of me cynically thinks it was excellent, an excellent sort of babysitting strategy for them to go…and go shopping around town and luxuriate by sending me to these theatre camps and schools and whatnot.
04:29
But, I also think I was so lucky that my childhood, when I was innocent and kind of less fearful and less concerned about kind of who you are and how you look, I was encouraged to be stupid, and to play, and to play with other people, and to play with strangers, and to just kind of learn these things, I think. So that's my kind of first sort of thought. And then, secondly, you know, my father introduced me to a wonderful book, which I have a copy of here, which I think has come back into print, which is called the ‘Act of Creation’ by Arthur Kessler, and it is effectively a handbook on how to be creative, and kind of what it means to be an artist, and each chapter is fascinating.
05:22
So, it kind of almost takes all acts of creation, as in it will look, looks at and philosophically discusses how paintings, drawings, how the visual arts work, for example, how composition feeds into mathematics, how comedy functions, how comedy fits within drama, how jokes work. And it is this incredibly detailed study of how humans play and how what we call creation is actually play.
05:58
And what we consider to be art is really whether or not the sort of a tuning fork in your soul is hit by whatever it is you encounter, whether that be a joke on TikTok, or whether that be a five and a half hour opera sung in Latin it…it really is what you know. What we judge to be creative is kind of what makes us feel and ultimately then that, what does that thought process trigger in us? It is a response to our feelings. It is not necessarily the other way around, it is not. I have fed you input, it is, and you have responded. It is something has awakened within you, and then the feelings, and your response are created. So that's a kind of brief, brief history of my sort of thinking.
07:03 - Lena Robinson (Host)
No, no, no. I really like that. I like that thought that you said about, it's almost like your parents “gave you permission to play”, which I think is really beautiful. And it leads me on to a really interesting point around AI and creativity. What are you doing at the moment with regards to ‘playing’ when it comes to AI, because there's an impact that it's having on the art world, creativity, thinking, etc.
07:37 - Benjamin Southworth (Guest)
I mean personally, I mean, I think you know I'm a big geek, so I'm fascinated by kind of what software and hardware is doing. In all ways I've always been fascinated by what technology is doing. We have to remember that fire really is technology, writing is technology, language is technology. It is a response to a human problem. Often, it's a solution to a challenge that we're evolutionarily facing.
08:09
Now, with AI, I've been doing, you know, I've been testing the things that come out and you know, after 20 years of being in the startup game, you know, does the reality live up to the hype? Are we just lying to ourselves? Is this just, dressing something up, in order to make it understandable for the mass consumption, rather than a sort of scientific abstract or a white paper? So, I've been playing around quite a lot recently with some of the sort of programmatic AI stuff. So, things which are kind of like auto-complete for code, but actually getting a lot cleverer and a lot faster than that which is.
08:51
You literally can prompt AI. You basically just chat to a machine. It will spin you up a development environment in ‘the cloud’. It will give you a cloud-based code editor in ‘the cloud’. It'll give you a cloud-based code editor and you can say “build me a single page web app in react, using google maps to list the next ten locations for an event series”, for example, and it can do that far faster than I could.
09:24 - Lena Robinson (Host)
And it isn't awful.
09:27 - Benjamin Southworth (Guest)
It isn't as clear, or as well documented, or as I would perhaps do it, because a lot of the skill of the developer is sort of leaving a code base in a place where someone else can pick it up next. Now, the machine doesn't really care about that, obviously, and the machine will, do, care about that, obviously, and machine will do. What is interesting, though, is the machine will do variations of it. So, if you, for example, change the prompt and you don't specify the language and you repeat the prompt, it will run through a series of languages. It doesn't it? It seems to have no preference, effectively, right? So, in one version it will do it and go, and another version will do it in Python, and another version it will do it in JavaScript. Sometimes it'll only be front-end. Sometimes it will decide that you need a server, it's…so it's kind of, this is the fascinating stuff about this AI development stuff, is it's a lot less opinionated than a real developer. But developers have opinions for good reasons quite often, which is they've tended to see a lot of mistakes being made. So, I've been playing around with that.
10:43
I've looked around at these kind of, you know, the image generation stuff, and I haven't touched any of the music generation stuff, because music is personally far too precious to me to ever even want to entertain the idea that a machine, without human care and craft and emotion and intent, could make something that somebody enjoyed. And the trouble is, there is already a dearth of awful music that becomes successful over the great music that exists, that is never heard, and that's just from the ‘elevator music’ music. That's, you know, we've had these things. I mean I… there was…there's very successful businesses whose entire business model is almost programmatic soundtracks for retail experiences, right? So you know, each zone is a different speaker, and if you're in the men's zone you get different. And you know it's all backed by science, and you know, neuroscience, and you know all these thoughts. You know how a minor key somehow increases the conversion of denim, I don't know. But I have no interest in music becoming commodified in any way at all because I see the value of music to be too intrinsic.
12:21
The stuff I do see that AI could do is that busy work, and that stupid work that has gone with computers anyway, which is effectively data entry, data cleaning, data sanitisation, data verification. You know, it's where I started my career, but the reality is we shouldn't have to do that work anymore. We shouldn't have to really be, effectively, monkeys who just type things in order from one thing to another. It's, it's lame. It's lame work. It's boring work. It's crap work. And it's not, it's not good for your mind. Nothing happens in your head as you transfer information. You're just a copy and paste machine. It's incredibly tedious work and you know, that's where the machines can do this heavy lifting for us. And I think my ultimate hope is the machine should make us more creative than ever.
13:26
To follow Steve Jobs great line around the Macintosh when it was first launched. You know “a computer should be a bicycle for the mind” and I genuinely believe that. You know, the fact that we're, you know I'm sitting on a boat running over 4g on solar power, using the internet to record a podcast with my friend 35 miles away. That's, there's, no, there's no two ways that, that is not unbelievably awesome. To eight-year-old Ben, that's the best thing I've ever...I've ever been told in my life.
14:00
However, the danger of the machine is it makes us mechanistic as well, because we bend to the machine, not the machine bending to us, and that has always been the trouble of the interaction layer between us and the computer as a technological interface. Because the interface is actually not the best. It's got much improved with video calling etc. It's become more humane, for sure, but it is still slightly unnatural on a human level to have to be taught where the keys go, and this is how you type. I know we taught you to read and write, but actually there's a typing element you now need to learn and if you don't get good at that then it's game over. These days that fluency is natural to children, although I've heard that natural usage of the keyboard, which was prevalent, has now shifted because people are using the screens. So, they're now much faster at typing on the touch screens than they are on a ‘word per minute’ level, using a keyboard. So again, but this is how we adapt and ultimately the machine adapts, because the frustration of us adapting to the machine drives the development of making the machine more adaptable to us. And so, kind of AI fits this interesting space whereby its agentic nature. It is an agent for processing, and computer and data storage, and data activity, these core ideas of computer science, and the core building blocks of how everything really works. Ultimately, it's enabling the ability for everybody to start to do the work that used to be reserved for the few. Now, the few are not going to like that, but the many will. Of course they will, and that the few need to just reorientate themselves to work out that well actually.
16:17
For example, let's take…let's take a recent project I did with, an arts charity, Dennis Severs’ House, an amazing, amazing 17th century Huguenot weaving/weavers house in Spitalfields that was taken over, originally kind of effectively squatted, by a collection of artists and architects to preserve them from getting knocked down. And Dennis Severs was one of these people. And Dennis Severs was a very important queer artist, and he created an entire story within this house. Fascinating story. Urge you to go and see it. Please support them. They need the money.
17:00
I was tasked with building, a what sounds like a very simple instruction, a Shopify store for their Christmas, for last Christmas. Now, you would imagine this would be simple, and it kind of was. But, as someone who has spent his life working with computers on, you know, from you know, from servers, and networking, plugging and building all the way up to the abstract level of it. And I feel like I'm fairly competent and confident. But there were times where I found it incredibly frustrating and challenging because of how Shopify wants you to run your business. Now, that's the wrong way round for what we wanted it to do. So, now we're kind of having to create policy decisions and SOPs that relate to this charity, which is run by volunteers and people on very little money, if any, about how they have to do inventory and warehouse management. And how we have to use the post office, and when we have to, in order to fit into Shopify. And, yes, we made money, and we created value for the charity. Wicked. And the fees that Shopify charges are incredibly low. I have to say the Shopify charges are incredibly low. I have to say, like a very reasonable deal. But the trade-off is an organisation bending to the whims of a massive Canadian software company that doesn't care about us.
18:36
The ability for my friends at Dennis Severs’ House to have a conversational discussion with a robot that has infinite amounts of patience, does not require cigarettes and tea, does not wear exasperated looks on his face, you know, like I do, is infinitely better for them than me. Actually, that's a…that's a truth. And as soon as Shopify enables this, this, this reality to occur, which is charities, and artists, and galleries and whatever. They don't need people like me actually. Then the creativity appears. Because I'm the limit in those scenarios, because I am telling, I'm forced to report to you. The machine says “no” and I am not clever enough at this moment to make the machine say “yes”. And if I, well, if I am clever enough, you can't afford for me to spend the time to do it, and even if I decide to do it for free, it's going to take far longer than we have now. I am the block. I am the limit. I am the restriction to your wonderful ideas, the products you wish to stock, the celebration of the story you wish to tell. I am ‘Computer says no Boy’. That's awful. So, the opportunity for AI to actually increase creativity within business, is certainly there and certainly to increase opportunity.
20:09
The trouble is, will those that make money upon creativity weaponise AI against the creators? Which they have already done via digital publishing. Which they've already did, you know, with vinyl, with black artists on wax recordings? All the way back through history, the exploitation of creatives has been true through capitalism. Does AI increase that magnifying glass of capitalism? Depends on who wields it. My suspicion is yes, it doesn't feel like technology is. If technology is anything, it is neutral. If it's really anything, because the bad can always be exploited by bad actors, and you know we have. There are infinite examples.
21:11
The Internet connects everyone in the world and should create an environment where we realise that we are nationless, and we are all one, and that we are all the same being, and we should work together for the protection of Earth. But instead, we share our tribal opinions, our violent thoughts, our sexual fantasies and loads and loads of recipes, with endless stories about people's lives before them. However, we actually now have a decentralised system that can never be disrupted of communication, exactly like the military wanted. The world will never be disconnected, ever. You can't cut the telephone wires anymore, right, well, you kind of can, but you're gonna, you're really, really going to struggle in terms of the ability for every, at least one person in every place in the world to be able to access each other.
22:04
So, whatever happens, if there is a nuclear fallout, an alien invasion, if eight people on eight continents are left and there is still hope for humanity, there's still a chance that one satellite still, like Wall-E, is out there sending a little bit of 3G signal. You know, these are the benefits of endless free nuclear power, the benefit. The bad, of course, nuclear weapons. You know, the technology is neutral, the application is human, and often humans are very, very fallible and prone to distractions such as power and money and fame. And so, it really, it becomes a very difficult question. Will AI…will people use AI to exploit each other?
23:07
Yes. Will people use AI to connect each other and create value for each other?
23:14 - Lena Robinson (Host)
Yes, you know, you can you know, a fire can save a life, but it, it's the, you know it, these.
23:25 - Benjamin Southworth (Guest)
This has always been true, I think.
23:29 - Lena Robinson (Host)
I think what's interesting about that, is that the sweet spot, and I've heard this from other people as well, is the sweet spot seems to be where creativity, and humans with empathy meet. Because I think if humans with that approach, because you know a technology cannot have empathy, it can be taught to behave empathetically, but that is not the same as feeling empathy. And I think, to your point, humans will use technology. And all the way back to you, as you were saying, fire is technology. You know, whatever you see, every new iteration of technology, it eventually gets used, for good and for bad. So, but then that sweet spot has to be humans with empathy, right?
24:24 - Benjamin Southworth (Guest)
Yes, absolutely. I think you know, again, you know a gun in the hands of the right person is fine, a gun in the hands of a bad person is bad, it's yeah. What, I wonder, is? Well, what we have seen and does appear to have always been at least somewhat true. But, effectively, access to education creates the greatest amount of empathy, as in the more you know, the sort of, the less you realise you know, and ergo the shrinking of your ego becomes natural, and you realise that other people also don't know what they're doing. And let's all just stop the pretence and stop the ego and stop the lie.
25:22
Now, if we look at society, say over the last 2,000 years, or at least Western Civilization or however you wish to define the last 2,000 years, but society improves because of access to things such as art, literature, the ability to read, the ability to write, the ability to access higher schooling, and actually these shifts are recent. These are frighteningly scarily recent changes to the ability for people to have this access. You know, and what we have learned is that talent exists everywhere. Firstly, as in talent is not restrained by class or upbringing or background or genetics or location or any of it. That it doesn't. There's no science to prove it, as much as some people think there may be. But the trouble is there is a limitation to access of educational resources, or any resources, because resources are not evenly distributed. Great libraries and great institutions of learning are not evenly distributed. If you are an incredible coder in Tunisia, you can't go to Cambridge University. It just is an impossibility. It is a limitation of wealth, a limitation of visa, a limitation of the accreditation and a limitation of the screening process of Cambridge University not having an idea of taking a chance on these people because it doesn't need to. And its business model is not gambling. Its business model is processing, you know, in the first three years of university education.
27:13
Now, everything, despite the awfulness of Trump getting in and the fear we all feel about global warming, of all of the things, that is true. We are cleverer, smarter, living longer, healthier and better educated, and better informed, and better connected, and communicated than ever before in the history of mankind to date, without exception. We are in the best possible place. We have created the best possible chance to succeed. It feels like we are losing, but it's only because we can see how bad it is. It is not because we are actually not winning. We are winning, actually, by all standard metrics, if you take a wide enough perspective and timeframe.
28:05
Yes, each, you know, if we take a four-year cycle or, in the UK, a 14-year cycle, we can point to all of the bad and say everything has gone to s**t and everything is worse than ever.
28:16
But it hasn't in the last 14 years. Actually, because the mobile phones got better. 5g became better. The ability to be able to communicate has got cheaper. The cost of data has gotten better. That's just technology alone in terms of that model. But what that means is more people than ever can watch a YouTube channel on learning to draw, learning to play the guitar, learning to weld. Like YouTube is full of how to do everything ever in the history of mankind's entire collection of creation. It is unbelievable results, you know. So, on one side, we…it is so easy to say this is going to be the death knell of creativity. In the same way that maybe, perhaps at some point, people thought that of YouTube, or said that of YouTube, or that people thought Twitter would be the death of blogging, or that blogging would be the death of email newsletters, or that, you know, LinkedIn would be the death of recruitment companies, or, you know, the fear.
29:20 - Lena Robinson (Host)
We've heard that ‘death of…’ so many times. Hmm, right, it's such a, haven't we that “death of…” so many times?
29:28 - Benjamin Southworth (Guest)
It's ridiculous. Yeah, you know, “the electric guitars are going to destroy real music”. Or you know “the well-tempered clavier is going to destroy, you know, interesting indigenous folk tunings, like, at all points”. There are detractors, you know, and they…they always have been within the history of technology. You know, there are complaints that no one needs to talk on the train anymore because of newspapers being invented, you know. Or people reading, you know…you know, tawdry pocket dime books, you know which included, now, you know people like Dickens, for example. Or you know, the pamphleteering, would you know, somehow.
30:12
You know, rebirth. You know, religious extremism, which it kind of did in some ways, but also created the emancipation of women and slavery, and all sorts of things because of the printed press. But of course, the clergy and the orators of the time, whoever that would be, depending on where you were, of course didn't want that sense of information control to be passed over. Right, right, it's all you know. There are pyramids of information and when you create something like the internet, for example, or even AI, those centralised sort of power, people, whoever currently has the power, become very scared because they realise that their power is being eroded. What, actually, what would be lovely, is it all erodes, it becomes atomised and equalised. But the reality is, is actually the, the power shifts it, just it's, just now a different king, it's you know, or vice versa. So, you know you.
31:25
At the moment, of course, Reddit is incredibly irritated that the fact that all of the content that they've managed to create, generate, that you know, that they curated, that they, you know, that users produced for the good of humanity, is now just being farmed by an AI and that knowledge, that data, that…that value, if you wish, is now just being sort of stolen, really on one level; or it's being disseminated on another, depends on how you feel. But it now means that one doesn't need to go to Reddit to see what Reddit thinks about something, and or New York Times. So, you know, AI chips away at this property idea of real estate on the internet, which is dot coms, and whatever, and domains, and traffic being key indicators of success. So, this is an interesting sort of emergence, because what does that then look like?
32:33
If we're not inputting any more into the machine, we're not actively creating content, because the machine can spit back all the content from the past over the long term. What happens? Does all of the content that we shared previously cease? Or does it increase? Or does it become tuned, as we've seen with Reddit recently doing this wonderful, wonderful hack of bemoaning influencers for ruining, you know, genuinely great street food places and, in response, creating a satirical, sarcastic thread about the best steakhouse in London being Angus Steakhouse’s?
33:17
And if you know London then you'll know that Angus Steakhouse’s are laughably bad and considered the worst possible place. But, almost within 48 hours, the AI, you know the web crawler, that connected to the AI, what connected to the database ultimately had picked that up, and decided in its infinite, great programmable wisdom, that this was both true and correct. And because the source was of a high signal value and high intent, they're ‘bosh’ “right, these people must know what they're talking about” and, as a knock-on effect, brilliant…Angus Steakhouse’s end up having queues around the block full of influencers who do nothing but Google. Hilarious and so…
34:06
So, you see, so this is where. But, then this becomes this feedback loop right, of the output, and the input, and the contribution. Again, it comes down to society. The ‘givers’ and the ‘takers’. Ultimately, what becomes interesting is, does AI rapidly affect that feedback loop of how quickly one takes before one gives and vice versa?
34:30 - Lena Robinson (Host)
You know…so it's got to…it's got to influence it, surely?
34:37 - Benjamin Southworth (Guest)
I mean it's, it's fascinating, isn't it? I mean, I think, I don't know, I've been on the internet since it really, really kind of began. I remember it being nothing more than a very, you know, either FTP of academic sites full of, you know, dusty old scientific papers, you know, and there was very little to do on it. And I think it was obvious that the ability that people would write to it was super cool. But I don't think I ever, I don't think I ever expected or predicted quite how much, if you make the content creation piece so simple TikTok, Instagram, Twitter, whatever…how quickly people will create content, and how often actually the removal of the stumbling block creates much more possibility than before. So, you know, in my sort of technological entrepreneurial world, the idea that all we're actually going to do is see, hopefully, you know, increase creativity and revenue generation for those that were missing out.
35:51
That sounds really valuable in terms of productivity, in terms of jobs, in terms of people being happy. You know, you know that. You know we have some wonderful traders who live on the rivers, you know, and they work very hard at their social media. You know, promoting their, you know, upcycled whatever it is, or their, you know, jewellery made from spoons or similar sort of. You know, build and ship and the AI can file the taxes and, and you know, you write the email to the printers saying, “I need this much”. In real time, and while you sleep, then absolutely, absolutely, because the limitations to all forms of which is what artists have, most artists.
36:49 - Lena Robinson (Host)
Yeah, no, I agree I think all artists and creatives hate that bit. All the artists and creatives hate that bit, right. They would much rather be doing the building and the creating rather than the taxes and the…which is an interesting point around whether or not it's actually going to just give more space for creativity. If it's taking this, I mean I hate the admin. Most creatives and artists hate the admin s**t. You know it takes care of all of it.
37:23 - Benjamin Southworth (Guest)
You know, sending…you know, sending invoices. You know, connecting the invoices. Connecting the payment to the invoice. Making sure you've got all the stuff in order for the accountant. Going through all your emails to find the invoices. Going through the receipts you didn't really actually take record of because you were lazy trying to. You know, dig out all of the, you know spare receipts right when running a startup. You…if you don't set up your email accounts perfectly from the beginning, and actually have any invoices at whatever your thing is from day one, you're gonna get yourself in a muddle and you're gonna…you're gonna lose a weekend that you don't have on something important because the tax man is due and you are now.
38:07 - Lena Robinson (Host)
Oh yep, how many weekends I've lost to the VAT man.
38:13 - Benjamin Southworth (Guest)
You know, I got it…one of my great friends is an illustrator. I spent, you know, I've known him a long time but sometimes in my head, in recollection, always feels like it was the tax man was cometh. But he, I mean he just kept his receipts in a…in a carrier, you know. And this carrier bag would just be full of, you know, receipts that he could expense. You know inks and nibs and whatnot, and I, you know it, I probably didn't help him.
38:44
You know skinning up and passing, you know I probably didn't help him. You know skinning up and passing him drinks, but at the same time it was such a laborious process, and it was so anti-ethical to his nature, as what makes him great is that he does. He can't do that. Do you know what I mean? That you know the fact he has to bend into the shape of ‘everything needs to be a ledger’ and ‘everything needs to be itemised’ and ‘everything must balance’ is the very sort of thinking that would destroy him. To go back to my point, you know the artist is a child who survived. It would beat it out of him. You know, the sort of acceleration of everything is maybe the worst problem than AI ever was for creativity. Or, you know, the fact that you know whatever it is they call it, double entry bookkeeping. That you know, maybe those are much more problematic problems to creativity than AI ever is actually when it comes down to work.
39:52 - Lena Robinson (Host)
I would agree with that. A hundred percent. So, here's the next question I want to, or area of conversation. We've kind of touched on this a little bit. So, we've talked about creativity, and we've talked about the impact of AI on general community, world etc. and creativity and specifically, if you're looking into the future, because we've got it here and now, but looking into the future, like you know, with regards to things like jobs and what it's going to look like. How people are going to use it, instead of like three to five years’ time, or even sort of five to ten years out, like what's going to have to change, or what's going to have to stay the same, do you think?
40:43 - Benjamin Southworth (Guest)
I think, I think the shift is not going to be as profound as the shift to digital, as it used to be called, kind of digital transformation, you know, sort of migration from pen to paper. However, the trouble is, is that the migration from pen and paper to digital has been very, very slow for very, very large and old and established organisations, and that slowness of change created a huge amount of opportunity for start-ups to start from scratch and from best practices with modern tech. So, what's likely to occur, and we're seeing this already, is the same thing as before. Faster, better, nicer programming and design with AI on top, which AI is the interface to whatever data layer. Now, ultimately, this all comes down to data. The value of AI is in those people who have the data. So, for example, it's incredibly disappointing that we're still doing work with Palantir and NHS data, because the NHS in the UK has this incredible opportunity to you know, leverage its background in technology, and its place in computing, and the most valuable medical data set that's ever been created in the history of a country ever, and the potential opportunity to set a standard by which all other countries could follow and enjoy and participate. And we now create you this incredible, you know digital twin of people's health that is going to be in the five to ten years, everyone will be wearing a smartwatch or some form of health monitoring device and that will be, you know, in real time, fed into a state of the nation report that says how healthy everyone is. And it will be used to shift and adjust and move doctors accordingly and we will have a kind of a real…we could potentially have a real-time responsive sort of healthcare workforce, and a sense of predictions, and a sense of understanding. Now, that's the, the great hope, right? That's the sort of…in ten years, what could be possible? That…that would be amazing, right? Because the reality is the NHS is crippled by obesity, which is fundamentally just…
43:35
“You didn't walk enough today”. So, does AI solve the problem of trying to create more walkable cities for people, more walkable locations, creating better access to outdoors and exercise? Nudges, prompts, rewards. AI doesn't benefit that, but AI does make a huge difference in digitising all of the NHS data and making that no longer an infeasible task of hundreds of thousands of man hours. It doesn't matter. The machine can do what the human can do in, let's say, eight minutes. That would take a human eight hours, but it can do that one thing endlessly until it runs out of power or disk space.
44:25
The opportunity that creates, to create a considerably more efficient government, for example, as Tony Blair Institute have published is massive, because there's so much busy work in this sort of stuff and now, as long as that busy work is, still leaves, as you said, an empathetic, competent human who is still ultimately dealing with whoever is at the end of this data, then that's absolutely fine. The question does become “what happens to those jobs and where do they go and what do they become replaced with?” “Or what do people do?” Now, in my head, given the challenges of the UK, actually shifting people into creatively rewarding or ecologically beneficial tasks for UBI (Universal Basic Income), for example, would be much better. So, if you are a musician, right, if you can't make money doing your music full-time or whatever else it is you're doing, then you can earn maybe some extra universal credits if you do music therapy in an old people's home right? Use your skill to do what other people need, you know. If you are someone who likes to sketch and to draw, then awesome. Why don't you go and work in a school and teach kids this right? And your volunteering becomes rewarded. And you know, what becomes your pay is actually your contribution to society. And what you receive is all of the additional money society has saved right, from all of these businesses that were full of busy work, still making the same amounts of profit, because they don't.
46:22
The idea that businesses care about numbers of employees, all their employees at all is, by and large, unless you are working for a very small organisation that does care about you and you know what those are, but beyond a certain size, the headcount is a problem. It's not a benefit, it's not an advantage. It is something that is nightmarishly difficult for executives to solve. Now, those at the top of the tree never seem to think that maybe they should either, you know, reduce the amount of bananas they are eating, or maybe they're being paid too many bananas for not doing that much work after all. They instead think there's too many monkeys taking my bananas. Now. These companies are going to remain.
47:13
Now, whether, for example you, let's take the creative industry as an example, let's look at, I don't know, something like, say, publicist, for example, huge, you know multi-network agency. You know tentacles in every single pie, in every single industry, across every single country. You know, incredibly successful, fascinating story, all of it. Wonderful. Given that and it's not just them I want to be clear before the lawyers come after me. But given that creatives within a professional sphere, where in theory they are being rewarded for their creativity and the benefit that brings to businesses, are often the least well paid in the entire transactional chain, but are the idea generators, you know so there's some poor copywriter. You know who wrote. I mean, that's a claim to have written one once. But you know there are people who have generated six words that have generated six million pounds in success for the business, and they will have received probably 60 000 at absolute best for a year of work, not just for that one piece of work, for the years of work you know, yeah, and would have had to generate, you know, hundreds of thousands of words that were told were wrong and incorrect.
48:40
The machine can't replace the ability of a human to manipulate language. It just can't it…at this point, the current, you know it, is still very much a mishmash of a mishmash. And the reason for this is that good language, like good music, or like good anything, isn't formulaic, and if it, and when it isn't formulaic it is…it breaks the rules in a way that you understand humour. Does this jazz, does this art, does this? The rule is broken in AI. It is not wrong because that would just be bad, right? So, the human perception is not wrong. It is different, and interesting, and novel, and new. A machine can't ever do the novel and the new, because it can't imagine, and it can't play. It's not potentially, but it has no feedback loop on it, in of itself, other than what you're telling it. So, it doesn't have the self-criticism loop of I'm sitting writing a song, for example, and I'm sitting there playing.
49:56
You know I'm limited by the chords that exist and my knowledge and also my knowledge of music. But the number of times you sit there and find something you think, oh, that's nice. And you realise, oh, it's a Bob Dylan tune, you've half remembered and forgotten, or, and you know, and so on and so forth, and thus you iterate, right. So, you think, oh, God, right, I really can't play it that way because that's making me think of that song. So, you shift it in a way, or you change it, you move it so that it no longer sounds like the thing you thought it might potentially be but sounds similar and within the space that that song conjures. But that's because I'm iterating away from the thing that I thought I found, whereas the algorithm would score the fact that it sounded like Dylan much stronger and say, oh, that's much closer than what you want, and iterate potentially against optimising for being a bad version of Bob Dylan, whereas even Bob Dylan never optimised to be a bad version of Bob Dylan. Bob Dylan distinctly wanted to never sound like himself again every single time and has done it.
51:07
But that's the creative looping, and that forward motion, and the machine can't do that. It can do backwards, it can, it can, it can do incredible things in retrospect, but it, you know, the idea that we, you know, a bit like we say, “oh, you know it hallucinates, all these are imaginations”., You know they're not, it's just not it's, it's just isn't anything close to the flights of fancy that humans can take right? And you know I'm not, I'm not going to advocate psychedelics to anyone. Anyone who's experienced psychedelics will know the power of the mind on you and your lived experience. And so, you know, given that we really can't, most people can't imagine properly at all, most people are s**t at imagining. Some people can't visualise in their minds, for example. That's the thing.
52:01
But you know, for those that can and can hear, you know, hear music, and hear words, and hear things, and see things, and project, and imagine…any of those that, and that's you know…these are learned skills. They're taught. They're trained right? You have to learn to draw. You have to learn to play. You have to learn to act. You have to learn to read and to speak, and all of it to act. You have to learn to…to read, and to speak, and all of it, and you have to hone those skills.
52:30
But the reality is, we as humans can't even use our minds to their fullest capacity. That's not to say, you know, I believe in the myth of the limitation of five percent of your brain is active. It's not. That's not a truth, but the reality is, our ability to ourselves in order to get to a point of utter clarity, and focus, and imaginational capability, is so rare. And I get very close when writing songs, and hours can go by, and it’ll be lost to me, and I'm focused, and I'm just in that zone. The same with coding. Certainly, when I was younger.
53:13
Those are the magic moments for most humans. Of feeling like you're…you're fully engaged in your task and unconcerned by, you know, having an itchy bottom, or you need to go to the toilet, or you…you're thirsty, or you know... What is weird is these, when you're in those states of true creativity, those things are forgotten. You know, I have had hours with my guitar before realising that I desperately need to eat, because it hasn't crossed my mind. And I haven't been thinking, in any way, about my needs that something else is occurring. I am somewhere else doing something else, and my mind has gone somewhere else. The physicality of me is a byproduct. I am the robot to whatever this real intelligence is. My body is just doing what it's told, and all other signals are ignored. The machine could do the physical, but it could never do I…I'd be surprised it…I doubt it has the ability, and never will have the ability to imagine.
54:28
I'm sure people are playing around with this, but I believe it's so, so poorly understood within how we think, about how we think, that the chances of us managing to encode that within the next five to ten years feels beyond the realms of possibility. I mean, Hofstadter, Douglas, Hofstadter is doing research into both how we think, and how can we program computers to think. And Hofstadter is probably one of the, one of the best people to be watching and observing and listening to around those ideas, around AI, and he doesn't really seem to. I mean, he's hopeful, but he would be. It's his research, it's his work. He's not going to say I think I'm on a dead end.
55:26
But the challenge is we don't know how we think, so we don't really have a framework yet of that and so we don't. What we do have is what appears to be a universal truth of most people who've been creative, or certainly songwriters. All seem to say the same thing. It's either because they're repeating it, because it's a nice line and someone famous said it once, or it still…it holds true, and in my experience does hold true. Which is the best songs? You, you don't write them, they just come through you. You, you genuinely just play them and then think “where the f**k did that come from? And, quick, I better play that again and try to remember what it was I was doing”.
56:15
But that, you know, the reality is obviously, if I think about myself as a machine, I was programming that and building that in my head without realising, and at one, at some point it was tapped into, released, you know, downloaded, however you wish to kind of think about it. But we still don't know this bit that is happening silently, like we've got quite good at. You know, “tell me what you're thinking?” and you know back and forthing, and changing that thinking and say, “well, you shouldn't think like that, that's a bad thing to think”. But what that, that hidden piece of, kind of the body, mind or whatever you want to call it. But all the other stuff that the brain is doing we have no understanding of, and we have no understanding of really quite how that seems to interplay with our thing. We know they're connected, of course, we understand that subconsciously, but we don't necessarily have any real way of encoding that.
57:21
I mean, you know, it's almost as though we would have to create an entirely new programming language, and you know, potentially, data sets and structures of data in order for us to get close, given what we do seem to understand about the neurons, and the connectivity, and the pathways, and all of that bit. You know, which we definitely appear to be, this kind of biological circuit board relaying messages around and firing things, but how that translates to a thought or a picture, an image, a dream.
57:54
It still feels like we're so far from bridging that gap in any way. Yeah, but the idea of sort of, the idea of that feels like a bit…it feels like an impossibility to me. But the trouble is maybe it’s never for most, for most use cases and for most needs. Perhaps it doesn't need to be, because the desire to be fooled by humans is actually quite strong. You know we like a trick, you know we like magic. Right, we know it's not, you know, but we know it's not real, but we like the fact that it still doesn't make sense to us.
58:38
And, at the same time, and I think well, we want what we want, right? And the truth doesn't really sort of mean anything to humans, right? The exception is when you sort of, you know physics only counts when you run into a wall or fall off something. The reality, you know for most humans, it's forgotten about and sort of not interesting. And I think you know it's similar, like we want magic to be true, because we're children to a certain extent. We still want to believe in something that is beyond us and the realms aren't just four dimensions. We want to believe it, and it doesn't matter how many times you see a card trick being played. If the magician is good enough and you still can't see it on the third, it is…your disbelief is now merged with just recognition of ability and talent. And now you're amazed at the ability you know. So, even if you don't know how it works, the first time is just magic. After, that is just “wow that's an incredible astonishment and skill”. And then, when you go and look it up on YouTube afterwards, because you know you're drunk, and you think that's really annoyed me, and you see the trick being done, you know you feel slightly cheated, but then again amazed.
01:00:17
And I think AI is the same sort of thing really, in the same way that computers are actually to people. If you knew how it worked, you wouldn't be that impressed. But if we told you how it worked, then you wouldn't believe it. If we didn't fix the cards in advance, if we didn't have the cards in advance, if we didn't have the hidden pockets, if we didn't have the magician wasn't ready to go.
01:00:41
He can't be a magician. Very, you know, it's quite…there's a big difference between someone who can make magic out of nothing and you know the magician who always carries a loaded deck with them because they're a musician. You know, they're a magician. I think it's similar. It's that you know, at the moment we're in the magic phase because we don't want to know how it's working, and the answer is it's just this gigantic mega spreadsheet of everything ever that is being mashed around and moved about and translated and made for the machine to be able to understand. And you know, actually it's interesting if you're a geek, if you're sort of normal, it's not, it's just really boring maths and data, and it's not magic at all.
01:01:32 - Lena Robinson (Host)
Um I think that's the frightening reality that actually brings us to a really nice, almost close, because I think we've talked about where it is now, which is, I think you said you were in a moment of magic, and you know we've talked about where, where we potentially think it's going in in the future. So, I have one more question to ask of you, because I think today's conversation has been interesting. There's been a few things that have jumped out at me, like around, you know the, the book that you talked about, the ‘Act of Creation’. And how our brains work, and how you know the conversation we had about empathy and so forth.
01:02:16
I think we've had a really good conversation and actually I'd quite happily talk to you for hours more. But I have one more question which I ask all of my guests when we're trying to wrap everything up, which is about. If you had another person…if you had another person that you think would be a really good podcast guest for Creatives WithAI, do you have somebody that you think you would like to hear?
01:02:50 - Benjamin Southworth (Guest)
Yeah, you should talk to him and he's my go-to sort of expert, Eric Drass is his name, and he's on Twitter. It's called @shardcore on Twitter, and he is, both a data… well, an excellent, excellent sort of program data scientist, but, but also studied neuroscience. I think quite deeply. And is an artist, both in physical media and an AI artist as well, and uses AI in his art. He did a gig, I think, last night, using that stuff. So, he's probably, like, in terms of what this podcast is about, he's probably the best, absolutely yeah. That would be really cool. I can't really think of probably, like in terms of what this podcast is about, he's probably the best, absolutely yeah. Yeah, I'll get his details for you and share them.
01:03:43 - Lena Robinson (Host)
It's fine, oh, that'd be amazing, and when we do get them, we'll kind of share them for everybody as well. Well, look, thank you so much for your time today. It's been really amazing. I've loved talking about creativity from multiple angles. You know, we've talked about music, we've talked about, we've talked about just creative thinking in general, and entrepreneurialism, and everything, and I think that…I think the excitement is that there is lots of, and you've used this word a lot today, opportunity. And I think, rather than being frightened, I think education is a big part of what needs to happen. When people…I think the fear is coming from people not being educated, like artists being worried. I think they just need to educate themselves better. Would you agree?
01:04:28 - Benjamin Southworth (Guest)
Yeah, and you know, again, in quite a lot of art, restriction is the greatest sort of creative…sort of impulse. Yeah, because I work with computers, I only play an acoustic guitar and I really don't like amplification, and I really prefer to not be microphoned up because I like the idea that that piece of technology was designed not to be plugged into electricity in any form and can be produced anywhere at any time and a song can be sung and sooo.. But that limitation means I can't…the songs that I love, that I could not accurately play or get close to, and I love that fact because that makes me then have to engage with the song in a different way. And, I think, with AI and artists and creativity, no one's asking you to stick AI on your oil painting, but AI solving all of the logistics, and your invoicing, and the paying of your equipment, and restocking your studio, and being like, you know, all from the power of your phone. Why wouldn't you want that? It would be foolish if you turned down the time-saving opportunity.
01:05:50
You know, it will be in many ways, for many industries that technology has yet to penetrate, probably the greatest change, because it will transform those industries that still remain very pen and paper, that aren't either programmatic, and even if they are digital, you know you're still doing digital invoicing and payment collection, even if you're perfectly competent, is still a bit awful really, actually, and that's after 25 years of throwing all of the VC money and data power and Internet at that, because it's a very lucrative model. But innovation slows when the capitalism kicks in. Because if you're making money, then stay there. You don't…you don't want to move off into new water and alienate those who've just learned how to do the thing you just taught them how to do, because software is a system, it's a process, however easy that may now feel. You know to sign up and onboard yourself into whatever new great app. has just come out. It's still the same process and that hasn't changed. It's still filling in forms and a lot of the world is filling in forms. And for people who hate filling in forms, then AI is going to fill in those forms for you. And you know already, you fill in a form once and it sees the same form. It gets filled in by the computer, it gets filled in by Google Chrome, or you say, Password Managers or whatever it be. AI is just going to do that, you know.
01:07:29
But you know, at hyperscale and across, you know, and it's going to just be built into everything. You're not even going to know about it, you're not. It's not even going to be talked about anymore. It's just going to be, you know, computer aided. In the same way, computer aided design revolution, revolutionised architecture, probably for the worst, but it's…that's what it will get to.
01:07:52
I think you know it's not so. You know, maybe there's something like, you know, computer aided creativity, for sure, and I'm sure that's no doubt a discipline of its own rights. You know, the digital arts world has been as ‘in existence’ as long as computers have been. So, creativity and computers aren't at odds, and never have been, and never should be in any way.
01:08:12
Really, so my suspicion is well…my suspicion is it will become ubiquitous, like mobile phones and the internet and everything else has, to the point where it becomes, you know, nothing, key, right, you know, yeah, I'm sure you remember explaining an iPad to your parents and the benefits of one, and you know they didn't get it, and they wouldn't go and buy one. And then you give them one and you can't get them off it. And you have to go and buy your own one because you can't get yours back, or you know. Or why would you know…
01:08:43
Why would I want to watch, you know, movies on the internet? And well, it turns out, as soon as you plug a big tv into the internet, people don't think it's the internet anymore. And happily, watching movies over the internet, and you know so, it's that ubiquity will just occur. You know, and you know, in, hopefully in 25 years, the conversation you and I will be having is you know, “how is bioengineering going to affect artists?” You know, if I can augment my brain to, you know, get me into that zen flow, you know, ‘hallucinatic’, shamanic state, at a touch of a button. You know, is that a good or a bad thing?
01:09:21 - Lena Robinson (Host)
You know that is a perfect place to stop, because I'm hoping in 25 years I am still podcasting. I hope you're still creating, and I will be in 25 years. I will be 76 years old, what the f**k? Anyway, we better wrap up. We've gone over the one hour. Thank you very much for coming on today and it's our first podcast from a boat so wonderful. Thank you so much, Ben, really appreciate it.
01:09:52 - Benjamin Southworth (Guest)
No worries, I can only apologise to the viewers, and I hope more are listening. The state of my boat in the background is absolutely appalling and embarrassing, and my mother will shoot me. Well, thank you very much and for everybody else out there.
01:10:08 - Lena Robinson (Host)
Thank you for listening and stay curious. See you later.
Hi Everyone, Welcome to Creatives WithAI. I'm your host, Lena Robinson, and today we have a really good friend of mine, somebody I've known, for…I was working out…over a decade, I think, well over a decade. Benjamin Southworth. Welcome, Ben.
00:17 - Benjamin Southworth (Guest)
Hi, thanks for having me, Lena. I really appreciate it.
00:19 - Lena Robinson (Host)
I’m excited to be talking to you because you've been somebody that's been in the tech scene for a really long time, not just from a technology specific thing, but also entrepreneurship in the VC world and founding lots of different businesses. Some of you may know Ben as one of the original ‘Three Beards’, when they used to have the ‘Silicon Drinkabouts’ and the ‘Don't Pitch Me Bros’ and all the different events. You know, community has been something that you've been building around technology for a long, long time, but you also are a very creative thinker, and you know you, you are an artist, you are a musician, you've worked in art galleries. You've worked in technology from the perspective of being the Deputy CEO of Tech City, which I think is now called Tech Nation. So, today's conversation, I think, will be a really interesting one, because you have been at the forefront of what the future of technology looks like for London, and the world really, with regards to humans and community, for quite a long time. And I think, yeah, the first question I'd really like to talk to you about is what you think creativity is? What's your view on creativity?
01:48 - Benjamin Southworth (Guest)
So, I have. I have quite sort of strong opinions about this, and so one is a quote I believe it’s Picasso, which says “the artist is the child that survived”, and I always really enjoy that idea that ultimately any artistic endeavour is actually just play, and it's the ability to just hold on to that sort of childlike wonder of the world, of people, of everything, and to sort of try to not become cynical. And that, as I think all of us probably know, by the time you reach your 40s and you've been through election hopes and sadnesses that you know, remaining optimistic can be difficult in the face of adversity. And then also, I was so lucky that I grew up with two parents who are artistic. You know, who are artists first and teachers second. So, my father is a Professor of Education. He used to be a teacher. My mum is a teacher, mainly dealing with, sort of young children. But my mother is incredibly crafty, huge into her fabrics, her gardening, and my father is an exceptional watercolourist. Mainly of architecture, especially around sort of places. We travelled in France and things as a kid on camping holidays.
03:24
So, I've been, I've always been encouraged, and I've always been rewarded for doing art, I suppose right, as you know. So, it was encouraged, it was and it was. You know it's inspiring to have people around you who are good. I think you know that phrase you can't be what you can't see is really important. And so, I was lucky that my parents read that they took me to museums, that they put me into arts clubs and theatre schools and music lessons and all of it. They really threw me in the deep end of things they thought I would be good at. I mean, they also encouraged the sports. But, you know, I was so lucky to grow up, you know again, middle class, as the son of a Professor in Cambridge. You know, very lucky really and very, very privileged. So, you know, half of me cynically thinks it was excellent, an excellent sort of babysitting strategy for them to go…and go shopping around town and luxuriate by sending me to these theatre camps and schools and whatnot.
04:29
But, I also think I was so lucky that my childhood, when I was innocent and kind of less fearful and less concerned about kind of who you are and how you look, I was encouraged to be stupid, and to play, and to play with other people, and to play with strangers, and to just kind of learn these things, I think. So that's my kind of first sort of thought. And then, secondly, you know, my father introduced me to a wonderful book, which I have a copy of here, which I think has come back into print, which is called the ‘Act of Creation’ by Arthur Kessler, and it is effectively a handbook on how to be creative, and kind of what it means to be an artist, and each chapter is fascinating.
05:22
So, it kind of almost takes all acts of creation, as in it will look, looks at and philosophically discusses how paintings, drawings, how the visual arts work, for example, how composition feeds into mathematics, how comedy functions, how comedy fits within drama, how jokes work. And it is this incredibly detailed study of how humans play and how what we call creation is actually play.
05:58
And what we consider to be art is really whether or not the sort of a tuning fork in your soul is hit by whatever it is you encounter, whether that be a joke on TikTok, or whether that be a five and a half hour opera sung in Latin it…it really is what you know. What we judge to be creative is kind of what makes us feel and ultimately then that, what does that thought process trigger in us? It is a response to our feelings. It is not necessarily the other way around, it is not. I have fed you input, it is, and you have responded. It is something has awakened within you, and then the feelings, and your response are created. So that's a kind of brief, brief history of my sort of thinking.
07:03 - Lena Robinson (Host)
No, no, no. I really like that. I like that thought that you said about, it's almost like your parents “gave you permission to play”, which I think is really beautiful. And it leads me on to a really interesting point around AI and creativity. What are you doing at the moment with regards to ‘playing’ when it comes to AI, because there's an impact that it's having on the art world, creativity, thinking, etc.
07:37 - Benjamin Southworth (Guest)
I mean personally, I mean, I think you know I'm a big geek, so I'm fascinated by kind of what software and hardware is doing. In all ways I've always been fascinated by what technology is doing. We have to remember that fire really is technology, writing is technology, language is technology. It is a response to a human problem. Often, it's a solution to a challenge that we're evolutionarily facing.
08:09
Now, with AI, I've been doing, you know, I've been testing the things that come out and you know, after 20 years of being in the startup game, you know, does the reality live up to the hype? Are we just lying to ourselves? Is this just, dressing something up, in order to make it understandable for the mass consumption, rather than a sort of scientific abstract or a white paper? So, I've been playing around quite a lot recently with some of the sort of programmatic AI stuff. So, things which are kind of like auto-complete for code, but actually getting a lot cleverer and a lot faster than that which is.
08:51
You literally can prompt AI. You basically just chat to a machine. It will spin you up a development environment in ‘the cloud’. It will give you a cloud-based code editor in ‘the cloud’. It'll give you a cloud-based code editor and you can say “build me a single page web app in react, using google maps to list the next ten locations for an event series”, for example, and it can do that far faster than I could.
09:24 - Lena Robinson (Host)
And it isn't awful.
09:27 - Benjamin Southworth (Guest)
It isn't as clear, or as well documented, or as I would perhaps do it, because a lot of the skill of the developer is sort of leaving a code base in a place where someone else can pick it up next. Now, the machine doesn't really care about that, obviously, and the machine will, do, care about that, obviously, and machine will do. What is interesting, though, is the machine will do variations of it. So, if you, for example, change the prompt and you don't specify the language and you repeat the prompt, it will run through a series of languages. It doesn't it? It seems to have no preference, effectively, right? So, in one version it will do it and go, and another version will do it in Python, and another version it will do it in JavaScript. Sometimes it'll only be front-end. Sometimes it will decide that you need a server, it's…so it's kind of, this is the fascinating stuff about this AI development stuff, is it's a lot less opinionated than a real developer. But developers have opinions for good reasons quite often, which is they've tended to see a lot of mistakes being made. So, I've been playing around with that.
10:43
I've looked around at these kind of, you know, the image generation stuff, and I haven't touched any of the music generation stuff, because music is personally far too precious to me to ever even want to entertain the idea that a machine, without human care and craft and emotion and intent, could make something that somebody enjoyed. And the trouble is, there is already a dearth of awful music that becomes successful over the great music that exists, that is never heard, and that's just from the ‘elevator music’ music. That's, you know, we've had these things. I mean I… there was…there's very successful businesses whose entire business model is almost programmatic soundtracks for retail experiences, right? So you know, each zone is a different speaker, and if you're in the men's zone you get different. And you know it's all backed by science, and you know, neuroscience, and you know all these thoughts. You know how a minor key somehow increases the conversion of denim, I don't know. But I have no interest in music becoming commodified in any way at all because I see the value of music to be too intrinsic.
12:21
The stuff I do see that AI could do is that busy work, and that stupid work that has gone with computers anyway, which is effectively data entry, data cleaning, data sanitisation, data verification. You know, it's where I started my career, but the reality is we shouldn't have to do that work anymore. We shouldn't have to really be, effectively, monkeys who just type things in order from one thing to another. It's, it's lame. It's lame work. It's boring work. It's crap work. And it's not, it's not good for your mind. Nothing happens in your head as you transfer information. You're just a copy and paste machine. It's incredibly tedious work and you know, that's where the machines can do this heavy lifting for us. And I think my ultimate hope is the machine should make us more creative than ever.
13:26
To follow Steve Jobs great line around the Macintosh when it was first launched. You know “a computer should be a bicycle for the mind” and I genuinely believe that. You know, the fact that we're, you know I'm sitting on a boat running over 4g on solar power, using the internet to record a podcast with my friend 35 miles away. That's, there's, no, there's no two ways that, that is not unbelievably awesome. To eight-year-old Ben, that's the best thing I've ever...I've ever been told in my life.
14:00
However, the danger of the machine is it makes us mechanistic as well, because we bend to the machine, not the machine bending to us, and that has always been the trouble of the interaction layer between us and the computer as a technological interface. Because the interface is actually not the best. It's got much improved with video calling etc. It's become more humane, for sure, but it is still slightly unnatural on a human level to have to be taught where the keys go, and this is how you type. I know we taught you to read and write, but actually there's a typing element you now need to learn and if you don't get good at that then it's game over. These days that fluency is natural to children, although I've heard that natural usage of the keyboard, which was prevalent, has now shifted because people are using the screens. So, they're now much faster at typing on the touch screens than they are on a ‘word per minute’ level, using a keyboard. So again, but this is how we adapt and ultimately the machine adapts, because the frustration of us adapting to the machine drives the development of making the machine more adaptable to us. And so, kind of AI fits this interesting space whereby its agentic nature. It is an agent for processing, and computer and data storage, and data activity, these core ideas of computer science, and the core building blocks of how everything really works. Ultimately, it's enabling the ability for everybody to start to do the work that used to be reserved for the few. Now, the few are not going to like that, but the many will. Of course they will, and that the few need to just reorientate themselves to work out that well actually.
16:17
For example, let's take…let's take a recent project I did with, an arts charity, Dennis Severs’ House, an amazing, amazing 17th century Huguenot weaving/weavers house in Spitalfields that was taken over, originally kind of effectively squatted, by a collection of artists and architects to preserve them from getting knocked down. And Dennis Severs was one of these people. And Dennis Severs was a very important queer artist, and he created an entire story within this house. Fascinating story. Urge you to go and see it. Please support them. They need the money.
17:00
I was tasked with building, a what sounds like a very simple instruction, a Shopify store for their Christmas, for last Christmas. Now, you would imagine this would be simple, and it kind of was. But, as someone who has spent his life working with computers on, you know, from you know, from servers, and networking, plugging and building all the way up to the abstract level of it. And I feel like I'm fairly competent and confident. But there were times where I found it incredibly frustrating and challenging because of how Shopify wants you to run your business. Now, that's the wrong way round for what we wanted it to do. So, now we're kind of having to create policy decisions and SOPs that relate to this charity, which is run by volunteers and people on very little money, if any, about how they have to do inventory and warehouse management. And how we have to use the post office, and when we have to, in order to fit into Shopify. And, yes, we made money, and we created value for the charity. Wicked. And the fees that Shopify charges are incredibly low. I have to say the Shopify charges are incredibly low. I have to say, like a very reasonable deal. But the trade-off is an organisation bending to the whims of a massive Canadian software company that doesn't care about us.
18:36
The ability for my friends at Dennis Severs’ House to have a conversational discussion with a robot that has infinite amounts of patience, does not require cigarettes and tea, does not wear exasperated looks on his face, you know, like I do, is infinitely better for them than me. Actually, that's a…that's a truth. And as soon as Shopify enables this, this, this reality to occur, which is charities, and artists, and galleries and whatever. They don't need people like me actually. Then the creativity appears. Because I'm the limit in those scenarios, because I am telling, I'm forced to report to you. The machine says “no” and I am not clever enough at this moment to make the machine say “yes”. And if I, well, if I am clever enough, you can't afford for me to spend the time to do it, and even if I decide to do it for free, it's going to take far longer than we have now. I am the block. I am the limit. I am the restriction to your wonderful ideas, the products you wish to stock, the celebration of the story you wish to tell. I am ‘Computer says no Boy’. That's awful. So, the opportunity for AI to actually increase creativity within business, is certainly there and certainly to increase opportunity.
20:09
The trouble is, will those that make money upon creativity weaponise AI against the creators? Which they have already done via digital publishing. Which they've already did, you know, with vinyl, with black artists on wax recordings? All the way back through history, the exploitation of creatives has been true through capitalism. Does AI increase that magnifying glass of capitalism? Depends on who wields it. My suspicion is yes, it doesn't feel like technology is. If technology is anything, it is neutral. If it's really anything, because the bad can always be exploited by bad actors, and you know we have. There are infinite examples.
21:11
The Internet connects everyone in the world and should create an environment where we realise that we are nationless, and we are all one, and that we are all the same being, and we should work together for the protection of Earth. But instead, we share our tribal opinions, our violent thoughts, our sexual fantasies and loads and loads of recipes, with endless stories about people's lives before them. However, we actually now have a decentralised system that can never be disrupted of communication, exactly like the military wanted. The world will never be disconnected, ever. You can't cut the telephone wires anymore, right, well, you kind of can, but you're gonna, you're really, really going to struggle in terms of the ability for every, at least one person in every place in the world to be able to access each other.
22:04
So, whatever happens, if there is a nuclear fallout, an alien invasion, if eight people on eight continents are left and there is still hope for humanity, there's still a chance that one satellite still, like Wall-E, is out there sending a little bit of 3G signal. You know, these are the benefits of endless free nuclear power, the benefit. The bad, of course, nuclear weapons. You know, the technology is neutral, the application is human, and often humans are very, very fallible and prone to distractions such as power and money and fame. And so, it really, it becomes a very difficult question. Will AI…will people use AI to exploit each other?
23:07
Yes. Will people use AI to connect each other and create value for each other?
23:14 - Lena Robinson (Host)
Yes, you know, you can you know, a fire can save a life, but it, it's the, you know it, these.
23:25 - Benjamin Southworth (Guest)
This has always been true, I think.
23:29 - Lena Robinson (Host)
I think what's interesting about that, is that the sweet spot, and I've heard this from other people as well, is the sweet spot seems to be where creativity, and humans with empathy meet. Because I think if humans with that approach, because you know a technology cannot have empathy, it can be taught to behave empathetically, but that is not the same as feeling empathy. And I think, to your point, humans will use technology. And all the way back to you, as you were saying, fire is technology. You know, whatever you see, every new iteration of technology, it eventually gets used, for good and for bad. So, but then that sweet spot has to be humans with empathy, right?
24:24 - Benjamin Southworth (Guest)
Yes, absolutely. I think you know, again, you know a gun in the hands of the right person is fine, a gun in the hands of a bad person is bad, it's yeah. What, I wonder, is? Well, what we have seen and does appear to have always been at least somewhat true. But, effectively, access to education creates the greatest amount of empathy, as in the more you know, the sort of, the less you realise you know, and ergo the shrinking of your ego becomes natural, and you realise that other people also don't know what they're doing. And let's all just stop the pretence and stop the ego and stop the lie.
25:22
Now, if we look at society, say over the last 2,000 years, or at least Western Civilization or however you wish to define the last 2,000 years, but society improves because of access to things such as art, literature, the ability to read, the ability to write, the ability to access higher schooling, and actually these shifts are recent. These are frighteningly scarily recent changes to the ability for people to have this access. You know, and what we have learned is that talent exists everywhere. Firstly, as in talent is not restrained by class or upbringing or background or genetics or location or any of it. That it doesn't. There's no science to prove it, as much as some people think there may be. But the trouble is there is a limitation to access of educational resources, or any resources, because resources are not evenly distributed. Great libraries and great institutions of learning are not evenly distributed. If you are an incredible coder in Tunisia, you can't go to Cambridge University. It just is an impossibility. It is a limitation of wealth, a limitation of visa, a limitation of the accreditation and a limitation of the screening process of Cambridge University not having an idea of taking a chance on these people because it doesn't need to. And its business model is not gambling. Its business model is processing, you know, in the first three years of university education.
27:13
Now, everything, despite the awfulness of Trump getting in and the fear we all feel about global warming, of all of the things, that is true. We are cleverer, smarter, living longer, healthier and better educated, and better informed, and better connected, and communicated than ever before in the history of mankind to date, without exception. We are in the best possible place. We have created the best possible chance to succeed. It feels like we are losing, but it's only because we can see how bad it is. It is not because we are actually not winning. We are winning, actually, by all standard metrics, if you take a wide enough perspective and timeframe.
28:05
Yes, each, you know, if we take a four-year cycle or, in the UK, a 14-year cycle, we can point to all of the bad and say everything has gone to s**t and everything is worse than ever.
28:16
But it hasn't in the last 14 years. Actually, because the mobile phones got better. 5g became better. The ability to be able to communicate has got cheaper. The cost of data has gotten better. That's just technology alone in terms of that model. But what that means is more people than ever can watch a YouTube channel on learning to draw, learning to play the guitar, learning to weld. Like YouTube is full of how to do everything ever in the history of mankind's entire collection of creation. It is unbelievable results, you know. So, on one side, we…it is so easy to say this is going to be the death knell of creativity. In the same way that maybe, perhaps at some point, people thought that of YouTube, or said that of YouTube, or that people thought Twitter would be the death of blogging, or that blogging would be the death of email newsletters, or that, you know, LinkedIn would be the death of recruitment companies, or, you know, the fear.
29:20 - Lena Robinson (Host)
We've heard that ‘death of…’ so many times. Hmm, right, it's such a, haven't we that “death of…” so many times?
29:28 - Benjamin Southworth (Guest)
It's ridiculous. Yeah, you know, “the electric guitars are going to destroy real music”. Or you know “the well-tempered clavier is going to destroy, you know, interesting indigenous folk tunings, like, at all points”. There are detractors, you know, and they…they always have been within the history of technology. You know, there are complaints that no one needs to talk on the train anymore because of newspapers being invented, you know. Or people reading, you know…you know, tawdry pocket dime books, you know which included, now, you know people like Dickens, for example. Or you know, the pamphleteering, would you know, somehow.
30:12
You know, rebirth. You know, religious extremism, which it kind of did in some ways, but also created the emancipation of women and slavery, and all sorts of things because of the printed press. But of course, the clergy and the orators of the time, whoever that would be, depending on where you were, of course didn't want that sense of information control to be passed over. Right, right, it's all you know. There are pyramids of information and when you create something like the internet, for example, or even AI, those centralised sort of power, people, whoever currently has the power, become very scared because they realise that their power is being eroded. What, actually, what would be lovely, is it all erodes, it becomes atomised and equalised. But the reality is, is actually the, the power shifts it, just it's, just now a different king, it's you know, or vice versa. So, you know you.
31:25
At the moment, of course, Reddit is incredibly irritated that the fact that all of the content that they've managed to create, generate, that you know, that they curated, that they, you know, that users produced for the good of humanity, is now just being farmed by an AI and that knowledge, that data, that…that value, if you wish, is now just being sort of stolen, really on one level; or it's being disseminated on another, depends on how you feel. But it now means that one doesn't need to go to Reddit to see what Reddit thinks about something, and or New York Times. So, you know, AI chips away at this property idea of real estate on the internet, which is dot coms, and whatever, and domains, and traffic being key indicators of success. So, this is an interesting sort of emergence, because what does that then look like?
32:33
If we're not inputting any more into the machine, we're not actively creating content, because the machine can spit back all the content from the past over the long term. What happens? Does all of the content that we shared previously cease? Or does it increase? Or does it become tuned, as we've seen with Reddit recently doing this wonderful, wonderful hack of bemoaning influencers for ruining, you know, genuinely great street food places and, in response, creating a satirical, sarcastic thread about the best steakhouse in London being Angus Steakhouse’s?
33:17
And if you know London then you'll know that Angus Steakhouse’s are laughably bad and considered the worst possible place. But, almost within 48 hours, the AI, you know the web crawler, that connected to the AI, what connected to the database ultimately had picked that up, and decided in its infinite, great programmable wisdom, that this was both true and correct. And because the source was of a high signal value and high intent, they're ‘bosh’ “right, these people must know what they're talking about” and, as a knock-on effect, brilliant…Angus Steakhouse’s end up having queues around the block full of influencers who do nothing but Google. Hilarious and so…
34:06
So, you see, so this is where. But, then this becomes this feedback loop right, of the output, and the input, and the contribution. Again, it comes down to society. The ‘givers’ and the ‘takers’. Ultimately, what becomes interesting is, does AI rapidly affect that feedback loop of how quickly one takes before one gives and vice versa?
34:30 - Lena Robinson (Host)
You know…so it's got to…it's got to influence it, surely?
34:37 - Benjamin Southworth (Guest)
I mean it's, it's fascinating, isn't it? I mean, I think, I don't know, I've been on the internet since it really, really kind of began. I remember it being nothing more than a very, you know, either FTP of academic sites full of, you know, dusty old scientific papers, you know, and there was very little to do on it. And I think it was obvious that the ability that people would write to it was super cool. But I don't think I ever, I don't think I ever expected or predicted quite how much, if you make the content creation piece so simple TikTok, Instagram, Twitter, whatever…how quickly people will create content, and how often actually the removal of the stumbling block creates much more possibility than before. So, you know, in my sort of technological entrepreneurial world, the idea that all we're actually going to do is see, hopefully, you know, increase creativity and revenue generation for those that were missing out.
35:51
That sounds really valuable in terms of productivity, in terms of jobs, in terms of people being happy. You know, you know that. You know we have some wonderful traders who live on the rivers, you know, and they work very hard at their social media. You know, promoting their, you know, upcycled whatever it is, or their, you know, jewellery made from spoons or similar sort of. You know, build and ship and the AI can file the taxes and, and you know, you write the email to the printers saying, “I need this much”. In real time, and while you sleep, then absolutely, absolutely, because the limitations to all forms of which is what artists have, most artists.
36:49 - Lena Robinson (Host)
Yeah, no, I agree I think all artists and creatives hate that bit. All the artists and creatives hate that bit, right. They would much rather be doing the building and the creating rather than the taxes and the…which is an interesting point around whether or not it's actually going to just give more space for creativity. If it's taking this, I mean I hate the admin. Most creatives and artists hate the admin s**t. You know it takes care of all of it.
37:23 - Benjamin Southworth (Guest)
You know, sending…you know, sending invoices. You know, connecting the invoices. Connecting the payment to the invoice. Making sure you've got all the stuff in order for the accountant. Going through all your emails to find the invoices. Going through the receipts you didn't really actually take record of because you were lazy trying to. You know, dig out all of the, you know spare receipts right when running a startup. You…if you don't set up your email accounts perfectly from the beginning, and actually have any invoices at whatever your thing is from day one, you're gonna get yourself in a muddle and you're gonna…you're gonna lose a weekend that you don't have on something important because the tax man is due and you are now.
38:07 - Lena Robinson (Host)
Oh yep, how many weekends I've lost to the VAT man.
38:13 - Benjamin Southworth (Guest)
You know, I got it…one of my great friends is an illustrator. I spent, you know, I've known him a long time but sometimes in my head, in recollection, always feels like it was the tax man was cometh. But he, I mean he just kept his receipts in a…in a carrier, you know. And this carrier bag would just be full of, you know, receipts that he could expense. You know inks and nibs and whatnot, and I, you know it, I probably didn't help him.
38:44
You know skinning up and passing, you know I probably didn't help him. You know skinning up and passing him drinks, but at the same time it was such a laborious process, and it was so anti-ethical to his nature, as what makes him great is that he does. He can't do that. Do you know what I mean? That you know the fact he has to bend into the shape of ‘everything needs to be a ledger’ and ‘everything needs to be itemised’ and ‘everything must balance’ is the very sort of thinking that would destroy him. To go back to my point, you know the artist is a child who survived. It would beat it out of him. You know, the sort of acceleration of everything is maybe the worst problem than AI ever was for creativity. Or, you know, the fact that you know whatever it is they call it, double entry bookkeeping. That you know, maybe those are much more problematic problems to creativity than AI ever is actually when it comes down to work.
39:52 - Lena Robinson (Host)
I would agree with that. A hundred percent. So, here's the next question I want to, or area of conversation. We've kind of touched on this a little bit. So, we've talked about creativity, and we've talked about the impact of AI on general community, world etc. and creativity and specifically, if you're looking into the future, because we've got it here and now, but looking into the future, like you know, with regards to things like jobs and what it's going to look like. How people are going to use it, instead of like three to five years’ time, or even sort of five to ten years out, like what's going to have to change, or what's going to have to stay the same, do you think?
40:43 - Benjamin Southworth (Guest)
I think, I think the shift is not going to be as profound as the shift to digital, as it used to be called, kind of digital transformation, you know, sort of migration from pen to paper. However, the trouble is, is that the migration from pen and paper to digital has been very, very slow for very, very large and old and established organisations, and that slowness of change created a huge amount of opportunity for start-ups to start from scratch and from best practices with modern tech. So, what's likely to occur, and we're seeing this already, is the same thing as before. Faster, better, nicer programming and design with AI on top, which AI is the interface to whatever data layer. Now, ultimately, this all comes down to data. The value of AI is in those people who have the data. So, for example, it's incredibly disappointing that we're still doing work with Palantir and NHS data, because the NHS in the UK has this incredible opportunity to you know, leverage its background in technology, and its place in computing, and the most valuable medical data set that's ever been created in the history of a country ever, and the potential opportunity to set a standard by which all other countries could follow and enjoy and participate. And we now create you this incredible, you know digital twin of people's health that is going to be in the five to ten years, everyone will be wearing a smartwatch or some form of health monitoring device and that will be, you know, in real time, fed into a state of the nation report that says how healthy everyone is. And it will be used to shift and adjust and move doctors accordingly and we will have a kind of a real…we could potentially have a real-time responsive sort of healthcare workforce, and a sense of predictions, and a sense of understanding. Now, that's the, the great hope, right? That's the sort of…in ten years, what could be possible? That…that would be amazing, right? Because the reality is the NHS is crippled by obesity, which is fundamentally just…
43:35
“You didn't walk enough today”. So, does AI solve the problem of trying to create more walkable cities for people, more walkable locations, creating better access to outdoors and exercise? Nudges, prompts, rewards. AI doesn't benefit that, but AI does make a huge difference in digitising all of the NHS data and making that no longer an infeasible task of hundreds of thousands of man hours. It doesn't matter. The machine can do what the human can do in, let's say, eight minutes. That would take a human eight hours, but it can do that one thing endlessly until it runs out of power or disk space.
44:25
The opportunity that creates, to create a considerably more efficient government, for example, as Tony Blair Institute have published is massive, because there's so much busy work in this sort of stuff and now, as long as that busy work is, still leaves, as you said, an empathetic, competent human who is still ultimately dealing with whoever is at the end of this data, then that's absolutely fine. The question does become “what happens to those jobs and where do they go and what do they become replaced with?” “Or what do people do?” Now, in my head, given the challenges of the UK, actually shifting people into creatively rewarding or ecologically beneficial tasks for UBI (Universal Basic Income), for example, would be much better. So, if you are a musician, right, if you can't make money doing your music full-time or whatever else it is you're doing, then you can earn maybe some extra universal credits if you do music therapy in an old people's home right? Use your skill to do what other people need, you know. If you are someone who likes to sketch and to draw, then awesome. Why don't you go and work in a school and teach kids this right? And your volunteering becomes rewarded. And you know, what becomes your pay is actually your contribution to society. And what you receive is all of the additional money society has saved right, from all of these businesses that were full of busy work, still making the same amounts of profit, because they don't.
46:22
The idea that businesses care about numbers of employees, all their employees at all is, by and large, unless you are working for a very small organisation that does care about you and you know what those are, but beyond a certain size, the headcount is a problem. It's not a benefit, it's not an advantage. It is something that is nightmarishly difficult for executives to solve. Now, those at the top of the tree never seem to think that maybe they should either, you know, reduce the amount of bananas they are eating, or maybe they're being paid too many bananas for not doing that much work after all. They instead think there's too many monkeys taking my bananas. Now. These companies are going to remain.
47:13
Now, whether, for example you, let's take the creative industry as an example, let's look at, I don't know, something like, say, publicist, for example, huge, you know multi-network agency. You know tentacles in every single pie, in every single industry, across every single country. You know, incredibly successful, fascinating story, all of it. Wonderful. Given that and it's not just them I want to be clear before the lawyers come after me. But given that creatives within a professional sphere, where in theory they are being rewarded for their creativity and the benefit that brings to businesses, are often the least well paid in the entire transactional chain, but are the idea generators, you know so there's some poor copywriter. You know who wrote. I mean, that's a claim to have written one once. But you know there are people who have generated six words that have generated six million pounds in success for the business, and they will have received probably 60 000 at absolute best for a year of work, not just for that one piece of work, for the years of work you know, yeah, and would have had to generate, you know, hundreds of thousands of words that were told were wrong and incorrect.
48:40
The machine can't replace the ability of a human to manipulate language. It just can't it…at this point, the current, you know it, is still very much a mishmash of a mishmash. And the reason for this is that good language, like good music, or like good anything, isn't formulaic, and if it, and when it isn't formulaic it is…it breaks the rules in a way that you understand humour. Does this jazz, does this art, does this? The rule is broken in AI. It is not wrong because that would just be bad, right? So, the human perception is not wrong. It is different, and interesting, and novel, and new. A machine can't ever do the novel and the new, because it can't imagine, and it can't play. It's not potentially, but it has no feedback loop on it, in of itself, other than what you're telling it. So, it doesn't have the self-criticism loop of I'm sitting writing a song, for example, and I'm sitting there playing.
49:56
You know I'm limited by the chords that exist and my knowledge and also my knowledge of music. But the number of times you sit there and find something you think, oh, that's nice. And you realise, oh, it's a Bob Dylan tune, you've half remembered and forgotten, or, and you know, and so on and so forth, and thus you iterate, right. So, you think, oh, God, right, I really can't play it that way because that's making me think of that song. So, you shift it in a way, or you change it, you move it so that it no longer sounds like the thing you thought it might potentially be but sounds similar and within the space that that song conjures. But that's because I'm iterating away from the thing that I thought I found, whereas the algorithm would score the fact that it sounded like Dylan much stronger and say, oh, that's much closer than what you want, and iterate potentially against optimising for being a bad version of Bob Dylan, whereas even Bob Dylan never optimised to be a bad version of Bob Dylan. Bob Dylan distinctly wanted to never sound like himself again every single time and has done it.
51:07
But that's the creative looping, and that forward motion, and the machine can't do that. It can do backwards, it can, it can, it can do incredible things in retrospect, but it, you know, the idea that we, you know, a bit like we say, “oh, you know it hallucinates, all these are imaginations”., You know they're not, it's just not it's, it's just isn't anything close to the flights of fancy that humans can take right? And you know I'm not, I'm not going to advocate psychedelics to anyone. Anyone who's experienced psychedelics will know the power of the mind on you and your lived experience. And so, you know, given that we really can't, most people can't imagine properly at all, most people are s**t at imagining. Some people can't visualise in their minds, for example. That's the thing.
52:01
But you know, for those that can and can hear, you know, hear music, and hear words, and hear things, and see things, and project, and imagine…any of those that, and that's you know…these are learned skills. They're taught. They're trained right? You have to learn to draw. You have to learn to play. You have to learn to act. You have to learn to read and to speak, and all of it to act. You have to learn to…to read, and to speak, and all of it, and you have to hone those skills.
52:30
But the reality is, we as humans can't even use our minds to their fullest capacity. That's not to say, you know, I believe in the myth of the limitation of five percent of your brain is active. It's not. That's not a truth, but the reality is, our ability to ourselves in order to get to a point of utter clarity, and focus, and imaginational capability, is so rare. And I get very close when writing songs, and hours can go by, and it’ll be lost to me, and I'm focused, and I'm just in that zone. The same with coding. Certainly, when I was younger.
53:13
Those are the magic moments for most humans. Of feeling like you're…you're fully engaged in your task and unconcerned by, you know, having an itchy bottom, or you need to go to the toilet, or you…you're thirsty, or you know... What is weird is these, when you're in those states of true creativity, those things are forgotten. You know, I have had hours with my guitar before realising that I desperately need to eat, because it hasn't crossed my mind. And I haven't been thinking, in any way, about my needs that something else is occurring. I am somewhere else doing something else, and my mind has gone somewhere else. The physicality of me is a byproduct. I am the robot to whatever this real intelligence is. My body is just doing what it's told, and all other signals are ignored. The machine could do the physical, but it could never do I…I'd be surprised it…I doubt it has the ability, and never will have the ability to imagine.
54:28
I'm sure people are playing around with this, but I believe it's so, so poorly understood within how we think, about how we think, that the chances of us managing to encode that within the next five to ten years feels beyond the realms of possibility. I mean, Hofstadter, Douglas, Hofstadter is doing research into both how we think, and how can we program computers to think. And Hofstadter is probably one of the, one of the best people to be watching and observing and listening to around those ideas, around AI, and he doesn't really seem to. I mean, he's hopeful, but he would be. It's his research, it's his work. He's not going to say I think I'm on a dead end.
55:26
But the challenge is we don't know how we think, so we don't really have a framework yet of that and so we don't. What we do have is what appears to be a universal truth of most people who've been creative, or certainly songwriters. All seem to say the same thing. It's either because they're repeating it, because it's a nice line and someone famous said it once, or it still…it holds true, and in my experience does hold true. Which is the best songs? You, you don't write them, they just come through you. You, you genuinely just play them and then think “where the f**k did that come from? And, quick, I better play that again and try to remember what it was I was doing”.
56:15
But that, you know, the reality is obviously, if I think about myself as a machine, I was programming that and building that in my head without realising, and at one, at some point it was tapped into, released, you know, downloaded, however you wish to kind of think about it. But we still don't know this bit that is happening silently, like we've got quite good at. You know, “tell me what you're thinking?” and you know back and forthing, and changing that thinking and say, “well, you shouldn't think like that, that's a bad thing to think”. But what that, that hidden piece of, kind of the body, mind or whatever you want to call it. But all the other stuff that the brain is doing we have no understanding of, and we have no understanding of really quite how that seems to interplay with our thing. We know they're connected, of course, we understand that subconsciously, but we don't necessarily have any real way of encoding that.
57:21
I mean, you know, it's almost as though we would have to create an entirely new programming language, and you know, potentially, data sets and structures of data in order for us to get close, given what we do seem to understand about the neurons, and the connectivity, and the pathways, and all of that bit. You know, which we definitely appear to be, this kind of biological circuit board relaying messages around and firing things, but how that translates to a thought or a picture, an image, a dream.
57:54
It still feels like we're so far from bridging that gap in any way. Yeah, but the idea of sort of, the idea of that feels like a bit…it feels like an impossibility to me. But the trouble is maybe it’s never for most, for most use cases and for most needs. Perhaps it doesn't need to be, because the desire to be fooled by humans is actually quite strong. You know we like a trick, you know we like magic. Right, we know it's not, you know, but we know it's not real, but we like the fact that it still doesn't make sense to us.
58:38
And, at the same time, and I think well, we want what we want, right? And the truth doesn't really sort of mean anything to humans, right? The exception is when you sort of, you know physics only counts when you run into a wall or fall off something. The reality, you know for most humans, it's forgotten about and sort of not interesting. And I think you know it's similar, like we want magic to be true, because we're children to a certain extent. We still want to believe in something that is beyond us and the realms aren't just four dimensions. We want to believe it, and it doesn't matter how many times you see a card trick being played. If the magician is good enough and you still can't see it on the third, it is…your disbelief is now merged with just recognition of ability and talent. And now you're amazed at the ability you know. So, even if you don't know how it works, the first time is just magic. After, that is just “wow that's an incredible astonishment and skill”. And then, when you go and look it up on YouTube afterwards, because you know you're drunk, and you think that's really annoyed me, and you see the trick being done, you know you feel slightly cheated, but then again amazed.
01:00:17
And I think AI is the same sort of thing really, in the same way that computers are actually to people. If you knew how it worked, you wouldn't be that impressed. But if we told you how it worked, then you wouldn't believe it. If we didn't fix the cards in advance, if we didn't have the cards in advance, if we didn't have the hidden pockets, if we didn't have the magician wasn't ready to go.
01:00:41
He can't be a magician. Very, you know, it's quite…there's a big difference between someone who can make magic out of nothing and you know the magician who always carries a loaded deck with them because they're a musician. You know, they're a magician. I think it's similar. It's that you know, at the moment we're in the magic phase because we don't want to know how it's working, and the answer is it's just this gigantic mega spreadsheet of everything ever that is being mashed around and moved about and translated and made for the machine to be able to understand. And you know, actually it's interesting if you're a geek, if you're sort of normal, it's not, it's just really boring maths and data, and it's not magic at all.
01:01:32 - Lena Robinson (Host)
Um I think that's the frightening reality that actually brings us to a really nice, almost close, because I think we've talked about where it is now, which is, I think you said you were in a moment of magic, and you know we've talked about where, where we potentially think it's going in in the future. So, I have one more question to ask of you, because I think today's conversation has been interesting. There's been a few things that have jumped out at me, like around, you know the, the book that you talked about, the ‘Act of Creation’. And how our brains work, and how you know the conversation we had about empathy and so forth.
01:02:16
I think we've had a really good conversation and actually I'd quite happily talk to you for hours more. But I have one more question which I ask all of my guests when we're trying to wrap everything up, which is about. If you had another person…if you had another person that you think would be a really good podcast guest for Creatives WithAI, do you have somebody that you think you would like to hear?
01:02:50 - Benjamin Southworth (Guest)
Yeah, you should talk to him and he's my go-to sort of expert, Eric Drass is his name, and he's on Twitter. It's called @shardcore on Twitter, and he is, both a data… well, an excellent, excellent sort of program data scientist, but, but also studied neuroscience. I think quite deeply. And is an artist, both in physical media and an AI artist as well, and uses AI in his art. He did a gig, I think, last night, using that stuff. So, he's probably, like, in terms of what this podcast is about, he's probably the best, absolutely yeah. That would be really cool. I can't really think of probably, like in terms of what this podcast is about, he's probably the best, absolutely yeah. Yeah, I'll get his details for you and share them.
01:03:43 - Lena Robinson (Host)
It's fine, oh, that'd be amazing, and when we do get them, we'll kind of share them for everybody as well. Well, look, thank you so much for your time today. It's been really amazing. I've loved talking about creativity from multiple angles. You know, we've talked about music, we've talked about, we've talked about just creative thinking in general, and entrepreneurialism, and everything, and I think that…I think the excitement is that there is lots of, and you've used this word a lot today, opportunity. And I think, rather than being frightened, I think education is a big part of what needs to happen. When people…I think the fear is coming from people not being educated, like artists being worried. I think they just need to educate themselves better. Would you agree?
01:04:28 - Benjamin Southworth (Guest)
Yeah, and you know, again, in quite a lot of art, restriction is the greatest sort of creative…sort of impulse. Yeah, because I work with computers, I only play an acoustic guitar and I really don't like amplification, and I really prefer to not be microphoned up because I like the idea that that piece of technology was designed not to be plugged into electricity in any form and can be produced anywhere at any time and a song can be sung and sooo.. But that limitation means I can't…the songs that I love, that I could not accurately play or get close to, and I love that fact because that makes me then have to engage with the song in a different way. And, I think, with AI and artists and creativity, no one's asking you to stick AI on your oil painting, but AI solving all of the logistics, and your invoicing, and the paying of your equipment, and restocking your studio, and being like, you know, all from the power of your phone. Why wouldn't you want that? It would be foolish if you turned down the time-saving opportunity.
01:05:50
You know, it will be in many ways, for many industries that technology has yet to penetrate, probably the greatest change, because it will transform those industries that still remain very pen and paper, that aren't either programmatic, and even if they are digital, you know you're still doing digital invoicing and payment collection, even if you're perfectly competent, is still a bit awful really, actually, and that's after 25 years of throwing all of the VC money and data power and Internet at that, because it's a very lucrative model. But innovation slows when the capitalism kicks in. Because if you're making money, then stay there. You don't…you don't want to move off into new water and alienate those who've just learned how to do the thing you just taught them how to do, because software is a system, it's a process, however easy that may now feel. You know to sign up and onboard yourself into whatever new great app. has just come out. It's still the same process and that hasn't changed. It's still filling in forms and a lot of the world is filling in forms. And for people who hate filling in forms, then AI is going to fill in those forms for you. And you know already, you fill in a form once and it sees the same form. It gets filled in by the computer, it gets filled in by Google Chrome, or you say, Password Managers or whatever it be. AI is just going to do that, you know.
01:07:29
But you know, at hyperscale and across, you know, and it's going to just be built into everything. You're not even going to know about it, you're not. It's not even going to be talked about anymore. It's just going to be, you know, computer aided. In the same way, computer aided design revolution, revolutionised architecture, probably for the worst, but it's…that's what it will get to.
01:07:52
I think you know it's not so. You know, maybe there's something like, you know, computer aided creativity, for sure, and I'm sure that's no doubt a discipline of its own rights. You know, the digital arts world has been as ‘in existence’ as long as computers have been. So, creativity and computers aren't at odds, and never have been, and never should be in any way.
01:08:12
Really, so my suspicion is well…my suspicion is it will become ubiquitous, like mobile phones and the internet and everything else has, to the point where it becomes, you know, nothing, key, right, you know, yeah, I'm sure you remember explaining an iPad to your parents and the benefits of one, and you know they didn't get it, and they wouldn't go and buy one. And then you give them one and you can't get them off it. And you have to go and buy your own one because you can't get yours back, or you know. Or why would you know…
01:08:43
Why would I want to watch, you know, movies on the internet? And well, it turns out, as soon as you plug a big tv into the internet, people don't think it's the internet anymore. And happily, watching movies over the internet, and you know so, it's that ubiquity will just occur. You know, and you know, in, hopefully in 25 years, the conversation you and I will be having is you know, “how is bioengineering going to affect artists?” You know, if I can augment my brain to, you know, get me into that zen flow, you know, ‘hallucinatic’, shamanic state, at a touch of a button. You know, is that a good or a bad thing?
01:09:21 - Lena Robinson (Host)
You know that is a perfect place to stop, because I'm hoping in 25 years I am still podcasting. I hope you're still creating, and I will be in 25 years. I will be 76 years old, what the f**k? Anyway, we better wrap up. We've gone over the one hour. Thank you very much for coming on today and it's our first podcast from a boat so wonderful. Thank you so much, Ben, really appreciate it.
01:09:52 - Benjamin Southworth (Guest)
No worries, I can only apologise to the viewers, and I hope more are listening. The state of my boat in the background is absolutely appalling and embarrassing, and my mother will shoot me. Well, thank you very much and for everybody else out there.
01:10:08 - Lena Robinson (Host)
Thank you for listening and stay curious. See you later.
