TRANSKRYPCJA 🔗
WEBVTT
00:00:21.394 --> 00:00:23.615
Welcome to another episode of Cloud
00:00:23.675 --> 00:00:24.155
Unplugged.
00:00:25.036 --> 00:00:29.617
So we have some topics obviously inspired
00:00:29.657 --> 00:00:30.417
by the news today.
00:00:33.396 --> 00:00:34.795
I'll kind of cover them in order that
00:00:34.816 --> 00:00:36.097
we're going to kind of discuss them.
00:00:36.116 --> 00:00:38.917
We've got the OpenAI's financial leak.
00:00:38.938 --> 00:00:41.258
So they were going to IPO and then
00:00:41.277 --> 00:00:42.698
there was a load of their finances that
00:00:42.719 --> 00:00:45.139
have managed to get leak prior to that.
00:00:46.320 --> 00:00:49.780
SpaceX buying Cursor for sixty bills.
00:00:51.402 --> 00:00:52.923
Quite epic amount of money.
00:00:54.543 --> 00:00:56.225
What's happening with cloud capacity at
00:00:56.265 --> 00:00:56.924
Microsoft?
00:00:58.165 --> 00:00:59.746
Because they're running GitHub in Amazon.
00:01:01.148 --> 00:01:04.590
And also a little bit of a flashback,
00:01:04.989 --> 00:01:05.850
Commodore is back.
00:01:07.251 --> 00:01:11.453
New Linux flip phone, basically.
00:01:11.993 --> 00:01:14.376
So let's get into OpenAI and the
00:01:14.396 --> 00:01:15.055
financials.
00:01:15.096 --> 00:01:16.257
I don't know if you've seen this, Salman.
00:01:17.280 --> 00:01:18.700
um that they were kind of leaked i
00:01:18.861 --> 00:01:21.162
don't know who leaked them but it was
00:01:21.183 --> 00:01:24.245
a guy called ed citron who's a journalist
00:01:24.286 --> 00:01:26.608
he's kind of got his own podcast show
00:01:26.647 --> 00:01:29.390
but um they were basically given to him
00:01:29.489 --> 00:01:31.992
somehow i don't know how um and there
00:01:32.012 --> 00:01:33.774
was obviously the audits of the financials
00:01:34.715 --> 00:01:35.734
there weren't there to be met i think
00:01:35.754 --> 00:01:37.316
there were there were the audits for the
00:01:37.977 --> 00:01:40.099
um for the ipo um
00:01:41.109 --> 00:01:43.191
it showed interesting numbers of like
00:01:43.210 --> 00:01:45.953
basically money moving back and forth um
00:01:45.992 --> 00:01:47.634
as in like some of them revenue they
00:01:47.694 --> 00:01:50.355
made was kind of under revenue from
00:01:50.394 --> 00:01:54.197
investors um so that was a bit peculiar
00:01:54.998 --> 00:01:58.359
um they are spending an insane amount of
00:01:58.400 --> 00:02:01.240
money um i think the operating cost of
00:02:01.260 --> 00:02:02.001
like twenty
00:02:02.721 --> 00:02:05.706
point two billion um or twenty point nine
00:02:05.725 --> 00:02:08.550
two billion or something um is what their
00:02:08.569 --> 00:02:09.931
losses are at the moment but they're going
00:02:09.951 --> 00:02:12.574
to be spending more as well even though
00:02:12.594 --> 00:02:13.736
they're making a loss they're going to
00:02:13.757 --> 00:02:16.419
spend more on that loss for more capex
00:02:16.460 --> 00:02:17.981
expenditure on data centers
00:02:19.022 --> 00:02:21.765
um and then they've like increased their
00:02:21.786 --> 00:02:23.949
uh marketing spend by five times and
00:02:23.968 --> 00:02:26.872
that's like up to five point seven three
00:02:27.012 --> 00:02:31.396
billion pounds on marketing um that
00:02:31.417 --> 00:02:32.638
they're kind of spending at the moment
00:02:32.657 --> 00:02:34.019
which is kind of a little bit wild
00:02:35.039 --> 00:02:37.763
anyway what do you think to those numbers
00:02:38.725 --> 00:02:39.006
Wow.
00:02:39.287 --> 00:02:41.388
Are these numbers even real, John?
00:02:41.468 --> 00:02:42.028
What is this?
00:02:42.169 --> 00:02:44.691
Is this a hallucination?
00:02:44.812 --> 00:02:46.233
These are the real numbers.
00:02:47.114 --> 00:02:48.215
These are the real numbers.
00:02:49.804 --> 00:02:53.025
That kind of indicate, obviously,
00:02:53.044 --> 00:02:54.626
we spoke about in the last podcast episode
00:02:54.645 --> 00:02:55.466
about all the hype,
00:02:55.866 --> 00:02:57.026
but we don't mean necessarily the
00:02:57.067 --> 00:02:59.247
technology hype because the technology is
00:02:59.288 --> 00:02:59.647
sound,
00:02:59.668 --> 00:03:04.348
but obviously the valuation hype and I
00:03:04.449 --> 00:03:07.270
guess the mind share that they're all
00:03:07.289 --> 00:03:08.411
trying to get of everybody.
00:03:09.741 --> 00:03:15.466
by just talking about mythos and all the
00:03:16.046 --> 00:03:17.907
fables, it's there, it's not there.
00:03:19.109 --> 00:03:23.233
And so it's like a constant AI washing
00:03:23.513 --> 00:03:26.195
almost constantly in the media.
00:03:27.597 --> 00:03:28.998
But whilst all that's happening and all
00:03:29.019 --> 00:03:29.680
the hypes there,
00:03:29.740 --> 00:03:34.127
obviously the valuation is inflated whilst
00:03:34.167 --> 00:03:34.848
all that's happening.
00:03:34.888 --> 00:03:36.189
And then this is kind of the truth
00:03:36.250 --> 00:03:36.590
in the end,
00:03:36.650 --> 00:03:38.212
the numbers get leaked and then you really
00:03:38.293 --> 00:03:38.793
see, well,
00:03:38.993 --> 00:03:41.798
the demand isn't really there to warrant
00:03:41.818 --> 00:03:42.538
that valuation.
00:03:44.277 --> 00:03:46.558
I think sometimes we just have to be
00:03:46.579 --> 00:03:49.000
a little bit careful with the numbers that
00:03:49.060 --> 00:03:49.300
show.
00:03:49.340 --> 00:03:50.560
But again, it's a leak,
00:03:50.580 --> 00:03:52.161
so we don't know how accurate the leak
00:03:52.221 --> 00:03:52.401
is.
00:03:52.980 --> 00:03:53.981
Within the same leak,
00:03:54.942 --> 00:03:56.582
one of the numbers that came out was
00:03:56.603 --> 00:03:59.163
that the company is getting more efficient
00:03:59.243 --> 00:03:59.364
at
00:04:01.515 --> 00:04:02.295
hosting losses.
00:04:02.536 --> 00:04:04.377
So in twenty twenty four,
00:04:04.456 --> 00:04:06.957
they were they spent
Load full transcript
two dollars thirty
00:04:06.997 --> 00:04:10.080
seven cents for every dollar of revenue.
00:04:10.379 --> 00:04:11.659
So, you know, there's a lot of losses.
00:04:12.099 --> 00:04:13.140
But in twenty twenty five,
00:04:13.181 --> 00:04:16.422
they're down to one dollar sixty for every
00:04:16.482 --> 00:04:17.963
dollar of revenue.
00:04:18.062 --> 00:04:20.004
And twenty twenty six, the numbers again,
00:04:20.903 --> 00:04:22.204
you could choose if you want to believe
00:04:22.764 --> 00:04:25.165
it's not a dollar twenty for every dollar
00:04:25.185 --> 00:04:25.526
spent.
00:04:25.846 --> 00:04:28.427
But that's within a dollar twenty now.
00:04:30.290 --> 00:04:31.170
All companies, John,
00:04:31.190 --> 00:04:32.291
that we know when they start,
00:04:32.572 --> 00:04:33.932
they operate at a loss.
00:04:34.434 --> 00:04:35.514
Like Uber, when they started,
00:04:35.574 --> 00:04:36.516
they operated at a loss.
00:04:37.596 --> 00:04:39.117
So much so that even now,
00:04:39.197 --> 00:04:41.800
especially with these LLM companies,
00:04:42.060 --> 00:04:44.783
we talked about GPT for taking costing
00:04:44.803 --> 00:04:47.846
around hundred million to just train and
00:04:48.226 --> 00:04:48.987
make it available.
00:04:49.687 --> 00:04:52.411
so yes the numbers are out yes they
00:04:52.430 --> 00:04:53.812
show that the companies are operating at
00:04:53.851 --> 00:04:56.855
loss but so do most tech companies
00:04:57.055 --> 00:04:59.776
alphabet meta in the beginning they do
00:04:59.836 --> 00:05:04.482
operate a loss now the the the reason
00:05:04.502 --> 00:05:06.923
why they got leaked again if they got
00:05:06.963 --> 00:05:09.505
leaked perhaps just so you and i can
00:05:09.526 --> 00:05:11.108
talk about it on this podcast john
00:05:11.528 --> 00:05:13.050
or others can talk about on this podcast,
00:05:13.071 --> 00:05:14.814
just to get their name out even more.
00:05:16.358 --> 00:05:17.459
Even people ask the question,
00:05:17.500 --> 00:05:19.343
is this company of any value,
00:05:19.423 --> 00:05:20.846
of the valuation that's been given?
00:05:21.226 --> 00:05:23.370
Is good marketing for the company itself?
00:05:24.877 --> 00:05:27.918
So I don't know if seeing these numbers,
00:05:27.939 --> 00:05:30.641
the headline numbers that are presented,
00:05:30.761 --> 00:05:32.081
does it mean anything for us?
00:05:32.442 --> 00:05:33.043
Perhaps not.
00:05:33.423 --> 00:05:34.903
Do you think this is good marketing?
00:05:35.644 --> 00:05:36.504
I think it's good marketing.
00:05:36.725 --> 00:05:36.985
Really?
00:05:37.024 --> 00:05:40.247
You think bad numbers are good marketing?
00:05:40.286 --> 00:05:42.209
Because people will ask a question.
00:05:42.649 --> 00:05:43.730
What is Anthropic doing?
00:05:44.209 --> 00:05:46.011
What is Mistral doing?
00:05:46.050 --> 00:05:47.271
What are the other companies doing?
00:05:48.624 --> 00:05:52.528
so again is this bad i don't necessarily
00:05:52.567 --> 00:05:55.872
think this is bad is it spending money
00:05:56.252 --> 00:05:59.475
in the company it's the same history in
00:05:59.754 --> 00:06:01.237
in the history we've seen all companies
00:06:01.276 --> 00:06:03.559
lose money in the beginning okay how much
00:06:03.579 --> 00:06:05.081
are you losing is a lot in this
00:06:05.120 --> 00:06:07.062
case of course i don't know what the
00:06:07.083 --> 00:06:09.404
comparison is to to uber i don't
00:06:09.625 --> 00:06:11.326
necessarily think this is too much of an
00:06:11.387 --> 00:06:11.648
issue
00:06:12.678 --> 00:06:13.579
I don't think it's too much of an
00:06:13.639 --> 00:06:16.120
issue, but I think what it showed, though,
00:06:16.380 --> 00:06:18.343
is the money movement.
00:06:18.923 --> 00:06:19.423
So,
00:06:19.622 --> 00:06:22.584
nine percent of the revenue came from
00:06:22.644 --> 00:06:24.625
Microsoft and SoftBank,
00:06:25.266 --> 00:06:26.848
which are obviously investors in the
00:06:26.908 --> 00:06:27.348
company.
00:06:27.408 --> 00:06:27.507
So,
00:06:27.528 --> 00:06:29.870
you've got money moving between other
00:06:29.910 --> 00:06:31.630
companies that are also investors back
00:06:31.690 --> 00:06:33.391
into the company that they're invested in
00:06:33.411 --> 00:06:35.072
that then inflates, obviously,
00:06:35.093 --> 00:06:35.973
the amount of revenue.
00:06:35.992 --> 00:06:39.095
So, I think it's more things like that.
00:06:39.435 --> 00:06:39.696
And then...
00:06:40.899 --> 00:06:43.262
all of the r d money which you
00:06:43.302 --> 00:06:46.125
kind of expect that was kind of spent
00:06:46.206 --> 00:06:48.988
and i think then they've invested in
00:06:49.028 --> 00:06:51.252
microsoft obviously for their expenditure
00:06:51.271 --> 00:06:52.994
for the compute and things like that for
00:06:53.074 --> 00:06:55.297
open ai and that money's then kind of
00:06:55.336 --> 00:06:56.377
returning back so it's just
00:06:57.440 --> 00:06:59.903
It's just oddity of the flow of where
00:06:59.923 --> 00:07:01.725
all the capital is going back and forth.
00:07:01.766 --> 00:07:03.228
And then the revenue is kind of part
00:07:03.247 --> 00:07:05.310
of that capital, which is not normal.
00:07:05.391 --> 00:07:05.670
Obviously,
00:07:05.711 --> 00:07:07.252
revenue you'd expect to be based on the
00:07:07.293 --> 00:07:08.233
things you are selling,
00:07:08.293 --> 00:07:10.396
not selling them to the companies that are
00:07:10.416 --> 00:07:11.197
already invested with you.
00:07:11.218 --> 00:07:11.617
But yeah.
00:07:12.610 --> 00:07:14.512
But if they have invested in you,
00:07:14.572 --> 00:07:16.552
but if they're selling a service to
00:07:16.612 --> 00:07:19.254
Microsoft, for example, Microsoft is,
00:07:19.793 --> 00:07:20.033
you know,
00:07:20.074 --> 00:07:22.175
you can go into Azure and start using
00:07:22.295 --> 00:07:23.136
open AI
models.
00:07:23.656 --> 00:07:25.916
And company as an organization will also
00:07:26.016 --> 00:07:27.596
use the open AI models, right?
00:07:28.017 --> 00:07:29.117
If the revenue is that,
00:07:29.158 --> 00:07:31.139
if that's where the revenue is coming
00:07:31.819 --> 00:07:34.259
from, I think that's legit.
00:07:34.319 --> 00:07:36.060
No problem whatsoever, right?
00:07:36.440 --> 00:07:38.461
Because they're selling a service.
00:07:39.101 --> 00:07:40.122
I don't know about SoftBank, John.
00:07:41.716 --> 00:07:42.357
Okay, interesting.
00:07:42.377 --> 00:07:42.877
I think so.
00:07:42.918 --> 00:07:44.540
Maybe SoftBank is using their models.
00:07:45.279 --> 00:07:46.281
Maybe they are using.
00:07:46.300 --> 00:07:47.041
I have no idea.
00:07:47.401 --> 00:07:50.805
I knew Sam has got to you,
00:07:51.326 --> 00:07:51.625
hasn't it?
00:07:51.725 --> 00:07:52.346
I just knew it.
00:07:53.108 --> 00:07:53.928
I saw you.
00:07:54.069 --> 00:07:57.172
I just knew this was happening.
00:07:57.271 --> 00:07:57.973
I told you, John.
00:07:58.951 --> 00:08:00.031
It's just what I'd like everyone to know,
00:08:00.050 --> 00:08:02.872
that before we even started this podcast,
00:08:02.891 --> 00:08:05.252
someone had to move a life-size cutout of
00:08:05.293 --> 00:08:07.353
Sam from behind that he had,
00:08:07.814 --> 00:08:09.173
that he had to take out of shot
00:08:09.733 --> 00:08:11.194
of the video before we started this
00:08:11.235 --> 00:08:11.615
podcast.
00:08:11.654 --> 00:08:12.975
So, I mean,
00:08:13.194 --> 00:08:15.636
weather is independent and there isn't
00:08:15.755 --> 00:08:17.776
a... I'm not really sure, but...
00:08:19.586 --> 00:08:22.829
But John, just humor me for a minute.
00:08:23.309 --> 00:08:25.571
We've seen this money move around between
00:08:25.591 --> 00:08:27.012
the same companies a lot, right?
00:08:27.653 --> 00:08:29.254
NVIDIA investing in OpenAI,
00:08:29.314 --> 00:08:31.216
OpenAI investing in...
00:08:31.297 --> 00:08:32.557
This money is just moving around.
00:08:33.359 --> 00:08:34.179
At the end of the day,
00:08:34.740 --> 00:08:37.783
who's going to win?
00:08:37.802 --> 00:08:38.624
Who's going to win at the end of
00:08:38.644 --> 00:08:38.844
the day?
00:08:40.225 --> 00:08:40.346
Well,
00:08:40.365 --> 00:08:42.087
we all know it's going to be Anthropic,
00:08:42.128 --> 00:08:42.307
right?
00:08:43.413 --> 00:08:44.432
Yeah, I mean,
00:08:44.894 --> 00:08:46.293
OpenAI is also going to get its share.
00:08:46.413 --> 00:08:47.934
Anthropic is also going to get its share.
00:08:48.934 --> 00:08:50.274
NVIDIA is also going to get its share.
00:08:50.455 --> 00:08:53.535
Because the issue at the moment is, yes,
00:08:54.236 --> 00:08:55.456
LLM technology is improving.
00:08:55.937 --> 00:08:58.857
It doesn't need as much compute that it
00:08:58.898 --> 00:09:00.418
needed in twenty twenty four to be able
00:09:00.438 --> 00:09:01.499
to train and to be able to do
00:09:01.599 --> 00:09:01.938
inference.
00:09:02.278 --> 00:09:03.799
The chips are now getting designed
00:09:04.500 --> 00:09:07.302
specifically for training and inference
00:09:07.543 --> 00:09:08.845
for these LLM models.
00:09:08.865 --> 00:09:10.687
So it's definitely going to get a bit
00:09:10.726 --> 00:09:10.907
more.
00:09:11.287 --> 00:09:13.889
Yes, it still costs a lot of money,
00:09:14.931 --> 00:09:17.193
but the people who are going to lose
00:09:17.254 --> 00:09:18.033
out is
00:09:19.154 --> 00:09:21.277
that the reason why this cost is because
00:09:21.297 --> 00:09:22.717
this training takes a lot of money.
00:09:22.878 --> 00:09:25.380
But then all these companies are moving
00:09:26.280 --> 00:09:28.302
LMS to usage based fee.
00:09:28.322 --> 00:09:30.985
Now, instead of giving you credits,
00:09:31.947 --> 00:09:33.288
I think what's going to happen is the
00:09:33.408 --> 00:09:35.850
end users will end up paying a lot
00:09:35.889 --> 00:09:38.231
more for these services because the
00:09:38.292 --> 00:09:39.192
companies realize, okay,
00:09:40.234 --> 00:09:41.153
Operating at a loss,
00:09:41.413 --> 00:09:43.254
we need to adjust our prices to reflect
00:09:43.654 --> 00:09:44.654
how much it actually costs.
00:09:44.975 --> 00:09:46.655
So people who will lose out as well.
00:09:47.635 --> 00:09:48.956
I don't necessarily think that means
00:09:49.015 --> 00:09:49.456
losing out.
00:09:49.755 --> 00:09:51.616
It's the same model every company uses.
00:09:51.777 --> 00:09:52.956
Uber, when they came in,
00:09:52.996 --> 00:09:53.836
they were much cheaper.
00:09:54.158 --> 00:09:55.457
They used to give more money to the
00:09:55.477 --> 00:09:58.839
drivers, but now they don't.
00:09:59.619 --> 00:10:00.818
I think the difference in this, though,
00:10:00.899 --> 00:10:03.460
is just the amount of money
00:10:04.446 --> 00:10:05.768
we're talking about like this isn't like
00:10:06.489 --> 00:10:08.831
you know five hundred grand here or like
00:10:09.192 --> 00:10:10.613
a couple of million over there we're
00:10:10.633 --> 00:10:13.176
talking like billions like billions that
00:10:13.375 --> 00:10:15.057
most companies aren't even valued at this
00:10:15.097 --> 00:10:18.863
amount so most you could buy an insane
00:10:18.903 --> 00:10:21.144
number of companies for that amount of
00:10:21.206 --> 00:10:22.506
money that's kind of going that they're
00:10:22.547 --> 00:10:23.707
talking about here which is
00:10:24.773 --> 00:10:26.914
all kind of fictitious to a degree.
00:10:26.975 --> 00:10:27.995
So, you know,
00:10:28.054 --> 00:10:30.898
it's not like it's evidenced hugely on the
00:10:30.937 --> 00:10:31.677
capital expenditure.
00:10:31.697 --> 00:10:33.519
So that's the bit that's a little bit
00:10:33.578 --> 00:10:35.061
warped because if you look at it through
00:10:35.081 --> 00:10:36.981
a market lens, its
own market,
00:10:37.822 --> 00:10:40.043
it's all quite, you know, on the up.
00:10:40.625 --> 00:10:43.046
The taxi market or the car hiring market,
00:10:43.105 --> 00:10:43.606
less so,
00:10:43.687 --> 00:10:46.408
or the delivery market or delivery market,
00:10:46.448 --> 00:10:47.629
probably less comparatively,
00:10:47.990 --> 00:10:49.331
less so because it's more stabilised.
00:10:49.350 --> 00:10:50.392
But I do know what you mean.
00:10:50.432 --> 00:10:52.033
Obviously, companies that start
00:10:53.339 --> 00:10:56.701
have to obviously make a lot of
00:10:56.741 --> 00:10:59.104
expenditure to kind of grow their foothold
00:10:59.423 --> 00:11:00.304
in those markets.
00:11:00.345 --> 00:11:02.726
But they're spending five billion on
00:11:02.788 --> 00:11:04.849
marketing, five billion on marketing.
00:11:04.869 --> 00:11:07.211
That's the cost of most companies.
00:11:07.792 --> 00:11:09.634
And you're spending five billion on
00:11:09.693 --> 00:11:12.956
marketing in a market to then get how
00:11:12.976 --> 00:11:13.977
much are you going to get back from
00:11:14.038 --> 00:11:15.359
five billion expenditure?
00:11:15.379 --> 00:11:18.280
What's the ROI on five billion on your
00:11:19.292 --> 00:11:21.255
know token usage like what are you
00:11:21.296 --> 00:11:23.139
actually going to get for that what's the
00:11:23.158 --> 00:11:24.561
what's the business model that allows you
00:11:24.600 --> 00:11:26.303
to know you can if i spend five
00:11:26.344 --> 00:11:29.028
billion on marketing i'm gonna get how
00:11:29.068 --> 00:11:31.432
much back as an roi on that marketing
00:11:33.139 --> 00:11:34.541
Yeah, this is hard, right?
00:11:34.561 --> 00:11:36.763
Because five billion in marketing seems
00:11:36.783 --> 00:11:37.243
quite a lot.
00:11:37.644 --> 00:11:39.144
But then these companies are taking out
00:11:39.184 --> 00:11:42.748
Super Bowl ads and videos every day and
00:11:42.827 --> 00:11:43.568
whatever's coming up.
00:11:43.589 --> 00:11:43.828
But yeah,
00:11:43.889 --> 00:11:45.149
five billion in marketing definitely
00:11:45.169 --> 00:11:45.529
sounds a lot.
00:11:45.590 --> 00:11:46.591
But, John,
00:11:46.691 --> 00:11:47.812
these companies are in a different
00:11:47.851 --> 00:11:49.533
ballpark now.
00:11:49.754 --> 00:11:51.934
We're talking about companies with a lot
00:11:51.975 --> 00:11:52.475
of backing.
00:11:52.916 --> 00:11:55.158
So them spending five billion is like
00:11:56.178 --> 00:11:57.740
other companies spending five hundred K.
00:11:58.721 --> 00:11:58.900
So
00:12:00.722 --> 00:12:01.903
I don't know.
00:12:01.943 --> 00:12:02.784
AI has got to me, John.
00:12:03.065 --> 00:12:04.186
The numbers don't seem real,
00:12:04.525 --> 00:12:10.072
but I'm in for a ride.
00:12:12.274 --> 00:12:13.835
Because it can't be quantified at the
00:12:13.875 --> 00:12:16.057
moment because it's so new,
00:12:16.096 --> 00:12:17.259
we don't know how people are going to
00:12:17.278 --> 00:12:18.220
make their money back,
00:12:18.240 --> 00:12:19.240
what the services are going to be.
00:12:19.280 --> 00:12:20.841
Obviously, money has been made.
00:12:20.881 --> 00:12:21.842
They are making money.
00:12:21.863 --> 00:12:23.083
They said they've made €
00:12:24.378 --> 00:12:25.918
Oh, no, they had a ten billion target,
00:12:26.958 --> 00:12:28.259
but I think they made thirteen billion.
00:12:29.639 --> 00:12:31.019
So obviously they are making money,
00:12:31.239 --> 00:12:33.700
and that's a very high revenue.
00:12:35.399 --> 00:12:38.701
But it does feel a little bit on
00:12:39.860 --> 00:12:41.400
the hype comparative to kind of other
00:12:41.441 --> 00:12:41.900
markets,
00:12:44.142 --> 00:12:45.741
but a little bit mental to me that
00:12:45.782 --> 00:12:47.981
that amount of capital has just been spent
00:12:48.903 --> 00:12:50.102
with all the other things kind of going
00:12:50.163 --> 00:12:50.243
on.
00:12:51.706 --> 00:12:53.008
But I think, John,
00:12:53.107 --> 00:12:54.988
the hype is there because the market is
00:12:55.067 --> 00:12:56.969
way bigger than it was for Uber,
00:12:57.009 --> 00:12:57.970
let's just say, right?
00:12:57.990 --> 00:12:59.250
Let's take that as an example.
00:12:59.490 --> 00:12:59.750
True.
00:12:59.809 --> 00:13:02.270
And the market is way bigger than it
00:13:02.311 --> 00:13:03.052
was for cloud.
00:13:03.172 --> 00:13:04.993
Yes, we have companies, you know, okay,
00:13:05.013 --> 00:13:07.453
we can talk about the cloud era as
00:13:07.494 --> 00:13:09.073
well when it all became cloud native.
00:13:09.114 --> 00:13:11.794
Of course, everybody needed to use,
00:13:11.894 --> 00:13:12.395
not everybody,
00:13:12.716 --> 00:13:14.096
but companies who wanted to host the
00:13:14.115 --> 00:13:16.356
services, they will go tend towards cloud,
00:13:16.496 --> 00:13:16.657
right?
00:13:16.677 --> 00:13:18.317
So you had all these companies pop up.
00:13:18.717 --> 00:13:19.619
But with these models, John,
00:13:20.458 --> 00:13:21.539
Everybody can use a model.
00:13:21.940 --> 00:13:24.302
Anybody who has access to a device can
00:13:24.381 --> 00:13:24.861
use a model.
00:13:26.043 --> 00:13:27.283
I know people in healthcare,
00:13:27.344 --> 00:13:29.365
they use this model for various reasons,
00:13:29.424 --> 00:13:29.625
right?
00:13:30.105 --> 00:13:31.947
And I know people who are not even
00:13:32.346 --> 00:13:33.087
like barbers,
00:13:33.107 --> 00:13:35.249
they're using these models to come up with
00:13:35.269 --> 00:13:36.470
some designs and whatnot.
00:13:36.529 --> 00:13:37.890
Barbara or barbers?
00:13:38.451 --> 00:13:38.912
A person.
00:13:39.152 --> 00:13:39.631
Barbers.
00:13:39.832 --> 00:13:40.332
Barbers.
00:13:40.413 --> 00:13:42.594
All the barbers.
00:13:42.934 --> 00:13:45.616
All of the barbers.
00:13:46.559 --> 00:13:49.081
out there today, if
your name is Barbara,
00:13:49.841 --> 00:13:52.383
you are using the model.
00:13:53.503 --> 00:13:56.323
And people are spinning up their own
00:13:56.365 --> 00:13:56.884
websites.
00:13:56.924 --> 00:13:58.304
They're doing their own design for the
00:13:58.345 --> 00:14:00.225
logos and running their businesses.
00:14:00.807 --> 00:14:02.407
So I think the market is way bigger
00:14:02.447 --> 00:14:04.788
for this than it was for anything I
00:14:04.807 --> 00:14:06.109
have seen myself.
00:14:06.188 --> 00:14:06.389
Right.
00:14:06.649 --> 00:14:11.770
I'm just talking about my time.
00:14:12.292 --> 00:14:13.091
I hate to say this, John,
00:14:13.111 --> 00:14:13.912
but you're right.
00:14:14.032 --> 00:14:14.692
Air has got to me.
00:14:15.513 --> 00:14:17.514
has got to me but i think that's
00:14:17.533 --> 00:14:18.994
why the numbers are too big that's why
00:14:19.014 --> 00:14:20.414
they want to spend that five billion
00:14:20.875 --> 00:14:22.274
because they think they can get a bigger
00:14:22.315 --> 00:14:24.655
share of the market and the competition is
00:14:24.716 --> 00:14:26.736
tough against open ai because you're up
00:14:26.777 --> 00:14:29.476
against anthropic who really is leading
00:14:29.496 --> 00:14:31.937
the market right anthropic is leading the
00:14:31.977 --> 00:14:35.318
market and hence why you know people say
00:14:35.359 --> 00:14:38.059
why do they models got banned i'm not
00:14:38.080 --> 00:14:39.821
saying anything john but yeah maybe
00:14:39.841 --> 00:14:40.880
there's got something to do with it
00:14:42.065 --> 00:14:42.144
Yeah,
00:14:42.164 --> 00:14:43.566
I don't want you going in an anthropic
00:14:44.687 --> 00:14:46.428
whilst you're a big advocate for open AI.
00:14:47.110 --> 00:14:49.130
I'm not an advocate for open AI.
00:14:49.150 --> 00:14:50.633
I'm not an advocate for entropic.
00:14:51.092 --> 00:14:52.894
No, we get it.
00:14:52.934 --> 00:14:54.275
We know where your alliances lie.
00:14:54.296 --> 00:14:56.957
But yeah, I don't know.
00:14:56.977 --> 00:15:00.240
The B to C aspect of it all,
00:15:00.321 --> 00:15:02.263
I don't think most people are building
00:15:02.322 --> 00:15:04.504
websites and doing all these things.
00:15:04.565 --> 00:15:06.366
I think they're uploading pictures and
00:15:06.386 --> 00:15:07.408
then being like, can you...
00:15:08.486 --> 00:15:08.826
I don't know,
00:15:09.147 --> 00:15:12.990
put long hair on me or can you
00:15:13.029 --> 00:15:13.450
do this thing?
00:15:14.750 --> 00:15:18.231
So the cost, though, of all those tokens,
00:15:18.653 --> 00:15:20.153
the general consumer B to C at the
00:15:20.192 --> 00:15:21.774
moment that you're saying it's a bit like
00:15:21.793 --> 00:15:22.235
the Uber,
00:15:23.195 --> 00:15:24.596
how sustainable that's going to be on the
00:15:24.635 --> 00:15:24.995
capex.
00:15:25.035 --> 00:15:27.317
I don't even know whether you'll even make
00:15:27.356 --> 00:15:28.638
a return on that cost.
00:15:29.859 --> 00:15:30.958
How much are you going to price those
00:15:30.979 --> 00:15:32.559
plans at based on the cost of the
00:15:32.600 --> 00:15:33.860
capex to get your money back?
00:15:34.461 --> 00:15:35.361
What's the volume of
00:15:36.311 --> 00:15:37.892
consumption because the average person
00:15:37.912 --> 00:15:39.812
that's using at the moment it's affordable
00:15:39.852 --> 00:15:41.453
but when it doesn't become affordable
00:15:41.494 --> 00:15:43.054
obviously people can't use it and won't be
00:15:43.075 --> 00:15:44.274
able to have the money to use it
00:15:44.816 --> 00:15:45.956
also let's be honest they won't be able
00:15:45.975 --> 00:15:47.096
to use it because they don't have a
00:15:47.157 --> 00:15:49.057
job i'm joking because they've all been
00:15:49.118 --> 00:15:51.798
taken by ai so um that was being
00:15:51.839 --> 00:15:55.961
facetious jesus uh but yeah i still i
00:15:55.981 --> 00:15:57.182
still think it's i still think it's a
00:15:57.221 --> 00:15:59.123
little bit out there i don't i feel
00:15:59.143 --> 00:16:01.565
it's definitely yet there's a market for
00:16:01.625 --> 00:16:01.705
it
00:16:02.804 --> 00:16:04.226
yes, it's a little bit more universal,
00:16:04.267 --> 00:16:05.828
but the same as all websites need cloud
00:16:05.869 --> 00:16:07.892
computing to host, you know,
00:16:08.052 --> 00:16:09.634
isn't that much different for like new
00:16:09.673 --> 00:16:12.498
services needing models to do features.
00:16:12.597 --> 00:16:14.922
I don't think it's that like radically
00:16:14.961 --> 00:16:16.624
different necessarily, or like,
00:16:17.414 --> 00:16:19.197
some market that is that differential
00:16:19.236 --> 00:16:21.799
because obviously to run things Netflix to
00:16:21.840 --> 00:16:23.282
get those services that you've got
00:16:23.322 --> 00:16:25.904
delivery or require internet or require
00:16:25.945 --> 00:16:28.248
cloud you know you wouldn't say that those
00:16:28.307 --> 00:16:31.350
markets are like eight trillion dollars
00:16:31.410 --> 00:16:32.873
big and all these other things you know
00:16:33.013 --> 00:16:36.277
so it's seems a little bit inflated but
00:16:36.681 --> 00:16:39.043
No, but that's market for you, right,
00:16:39.082 --> 00:16:39.764
John?
00:16:40.804 --> 00:16:41.865
It runs on sentiment.
00:16:42.105 --> 00:16:44.427
The sentiment behind this technology is
00:16:44.467 --> 00:16:45.168
very bullish.
00:16:45.629 --> 00:16:46.210
That's the point.
00:16:46.769 --> 00:16:48.131
And that's why all these numbers seem
00:16:48.152 --> 00:16:48.652
conflated.
00:16:48.711 --> 00:16:50.433
But what are these numbers?
00:16:50.553 --> 00:16:51.215
Are they even real?
00:16:51.414 --> 00:16:51.995
I don't know, John.
00:16:52.436 --> 00:16:53.275
I'm just here for the ride.
00:16:53.316 --> 00:16:54.977
Well, we know where your sentiment
lies.
00:16:55.457 --> 00:16:56.438
That's for sure.
00:16:56.639 --> 00:17:01.783
So we move on to the SpaceX.
00:17:03.105 --> 00:17:06.527
purchase of Cursor as well,
00:17:06.606 --> 00:17:10.368
which was also a bit of a curveball,
00:17:10.388 --> 00:17:11.750
but quite smart, I think,
00:17:13.391 --> 00:17:15.592
really by Elon in kind of trying to
00:17:15.612 --> 00:17:18.834
get into the code area and to be
00:17:18.854 --> 00:17:20.555
kind of more of a competitor to OpenAI
00:17:20.694 --> 00:17:21.655
and Anthropic and others.
00:17:21.736 --> 00:17:23.737
So it does kind of make sense.
00:17:23.777 --> 00:17:26.778
But sixty billion, obviously, of stock,
00:17:28.420 --> 00:17:31.240
basic stock to acquire Cursor.
00:17:31.342 --> 00:17:32.041
Cursor being
00:17:33.915 --> 00:17:34.395
Obviously,
00:17:34.556 --> 00:17:36.538
they're for helping developers write code.
00:17:36.557 --> 00:17:37.438
They have their own model.
00:17:38.499 --> 00:17:43.983
They extended the IDE for VS Code.
00:17:45.263 --> 00:17:47.946
What did you think about that acquisition?
00:17:49.539 --> 00:17:50.740
Well, first of all, John,
00:17:51.339 --> 00:17:52.520
let's start with this.
00:17:52.560 --> 00:17:55.661
SpaceX IPO-ed last week, week before.
00:17:56.201 --> 00:17:58.582
SpaceX, one trillion, right?
00:17:58.602 --> 00:17:59.142
The IPO,
00:17:59.182 --> 00:18:01.482
the share price started at one sixty,
00:18:01.942 --> 00:18:03.103
one hundred and sixty dollars,
00:18:03.903 --> 00:18:07.243
and it shot up to two hundred within
00:18:07.284 --> 00:18:07.884
a day, right?
00:18:07.983 --> 00:18:08.683
Two hundred dollars.
00:18:09.124 --> 00:18:10.285
And now it's settled down to like one
00:18:10.305 --> 00:18:11.164
hundred and eighty to one hundred and
00:18:11.184 --> 00:18:12.904
eighty five.
00:18:12.944 --> 00:18:14.585
So before we even talk about Cursor,
00:18:14.986 --> 00:18:16.365
the IPO for SpaceX,
00:18:17.579 --> 00:18:19.941
was a success because it didn't go down
00:18:20.240 --> 00:18:24.462
below the IPO price.
00:18:25.663 --> 00:18:29.025
And SpaceX doesn't just include SpaceX
00:18:29.065 --> 00:18:29.424
itself.
00:18:29.765 --> 00:18:31.405
It also includes Starlink.
00:18:31.645 --> 00:18:35.208
It also includes XAI, John.
00:18:35.327 --> 00:18:37.449
So now some of these moves are,
00:18:37.489 --> 00:18:37.929
of course,
00:18:37.969 --> 00:18:42.090
being done to increase their footprint.
00:18:43.123 --> 00:18:47.284
Because we have a player, Anthropic,
00:18:47.403 --> 00:18:50.464
in development LLMs,
00:18:50.585 --> 00:18:51.625
and we also have OpenAI.
00:18:51.684 --> 00:18:52.884
But really, there's no other.
00:18:54.025 --> 00:18:57.925
So because now SpaceX have that capital,
00:18:59.185 --> 00:19:00.267
I think it's good for them.
00:19:00.846 --> 00:19:02.106
Well, that's what's happening.
00:19:02.126 --> 00:19:05.627
They've taken a bet on the billion dollars
00:19:05.667 --> 00:19:06.248
that you said, right?
00:19:07.304 --> 00:19:08.684
Now, Cursor, as we know,
00:19:08.785 --> 00:19:12.388
is a developer platform that people can
00:19:12.648 --> 00:19:14.691
use to use any models they like.
00:19:15.070 --> 00:19:17.232
And I think, John, at the moment,
00:19:17.513 --> 00:19:21.457
Cursor spends a bunch of their revenue to
00:19:21.537 --> 00:19:21.998
Anthropic.
00:19:22.057 --> 00:19:22.617
Is that correct?
00:19:22.699 --> 00:19:27.303
Like Anthropic do charge some of that
00:19:27.343 --> 00:19:29.785
because they use the Anthropic models?
00:19:29.825 --> 00:19:30.546
Claude, is that right?
00:19:31.326 --> 00:19:32.566
Oh, I didn't actually know that.
00:19:32.846 --> 00:19:33.606
I don't know.
00:19:33.626 --> 00:19:33.987
Is that right?
00:19:34.047 --> 00:19:34.707
I didn't know that bit.
00:19:34.787 --> 00:19:35.988
I know they have their own model called
00:19:36.028 --> 00:19:39.269
Composer or that was like what they built.
00:19:39.528 --> 00:19:42.490
Yeah, you can also use it.
00:19:42.509 --> 00:19:44.550
So I think the idea here is that
00:19:47.230 --> 00:19:48.832
SpaceX is going to I don't know if
00:19:48.852 --> 00:19:50.152
they have the boarded yet or is it
00:19:50.172 --> 00:19:51.593
still in process of buying?
00:19:52.032 --> 00:19:53.153
I thought they had bought it.
00:19:53.272 --> 00:19:56.454
Yeah, but I assume they had bought it.
00:19:59.019 --> 00:19:59.839
But yeah,
00:19:59.920 --> 00:20:08.570
I reckon it's to take down OpenAI,
00:20:08.711 --> 00:20:09.231
I think.
00:20:09.491 --> 00:20:10.673
I reckon he's going to be a gun
00:20:10.752 --> 00:20:12.536
for Sam Altman.
00:20:13.336 --> 00:20:15.138
Yeah, so...
00:20:15.875 --> 00:20:18.376
I think that's a fair point because Elon
00:20:18.396 --> 00:20:19.738
Musk and Sam Altman,
00:20:19.758 --> 00:20:21.638
they started OpenAI and there was a
00:20:21.679 --> 00:20:23.819
fallout and Elon Musk left OpenAI and
00:20:24.000 --> 00:20:25.621
there's been a bit of back and forth.
00:20:26.321 --> 00:20:29.604
But I think then we probably will see
00:20:29.644 --> 00:20:30.203
that Grok
00:20:31.334 --> 00:20:36.277
models will be competition for the models
00:20:36.297 --> 00:20:38.598
by OpenAI and Anthropic.
00:20:38.719 --> 00:20:41.660
Because right now, Grok models exist,
00:20:42.201 --> 00:20:43.662
but they don't have a massive user base.
00:20:43.701 --> 00:20:46.522
People who are using X use those models.
00:20:47.223 --> 00:20:48.545
Developers don't use those models.
00:20:49.817 --> 00:20:50.836
Where do developers live?
00:20:51.116 --> 00:20:52.538
Well, they live in Cursor at the moment,
00:20:52.577 --> 00:20:52.758
right?
00:20:52.857 --> 00:20:54.877
And that's one of the toolings people use
00:20:55.337 --> 00:20:56.038
for these models.
00:20:56.439 --> 00:20:59.179
So perhaps it is to capture that developer
00:20:59.199 --> 00:21:01.160
market just so they can.
00:21:01.539 --> 00:21:02.480
I think you're probably right.
00:21:02.839 --> 00:21:05.101
They're gunning for OpenAI and Anthropic,
00:21:05.740 --> 00:21:07.101
not just OpenAI, both of them,
00:21:07.540 --> 00:21:08.981
because this is a huge market.
00:21:09.701 --> 00:21:13.462
So they did quite well with the SpaceX
00:21:14.042 --> 00:21:16.203
IPO, but they probably will do way better.
00:21:18.065 --> 00:21:19.244
Now it's just marketplace.
00:21:19.565 --> 00:21:20.164
You're in the IPO.
00:21:20.184 --> 00:21:22.165
Now you have to keep making moves that
00:21:22.185 --> 00:21:24.426
the market is going to like and the
00:21:24.446 --> 00:21:25.567
sentiment is going to increase.
00:21:25.586 --> 00:21:26.207
It will get better.
00:21:26.567 --> 00:21:29.407
So I think this is a reasonable move
00:21:29.428 --> 00:21:30.469
for SpaceX to make.
00:21:30.909 --> 00:21:32.689
Now you're going to tell me that Elon
00:21:32.709 --> 00:21:33.369
has got me too.
00:21:33.690 --> 00:21:34.650
SpaceX has got me too.
00:21:35.029 --> 00:21:37.349
I'm just talking about from market
00:21:37.390 --> 00:21:37.990
sentiment, right?
00:21:38.030 --> 00:21:40.310
I'm not saying this is right or this
00:21:40.330 --> 00:21:40.651
is wrong.
00:21:40.671 --> 00:21:41.911
I'm just talking about a market sentiment.
00:21:42.351 --> 00:21:43.510
If you're running this company,
00:21:43.932 --> 00:21:46.011
you want to increase your market share and
00:21:46.031 --> 00:21:47.392
you want to increase your market cap,
00:21:47.612 --> 00:21:48.392
what would you do?
00:21:49.311 --> 00:21:50.893
And this is where the hot topic is.
00:21:50.913 --> 00:21:52.393
But you don't think though, I mean,
00:21:52.413 --> 00:21:53.032
obviously for him,
00:21:53.053 --> 00:21:53.772
it didn't really mean very much.
00:21:53.813 --> 00:21:55.413
I think it was like three percent equity
00:21:56.253 --> 00:21:57.973
essentially of SpaceX.
00:21:57.993 --> 00:21:59.513
That's kind of what he's given up to
00:21:59.534 --> 00:22:00.433
then acquire Cursor.
00:22:00.453 --> 00:22:01.454
So obviously it hasn't really cost
00:22:01.515 --> 00:22:03.055
anything necessarily per se.
00:22:05.469 --> 00:22:07.430
But Claude Code, I think, has like,
00:22:07.490 --> 00:22:09.751
forty-six percent of the market.
00:22:09.852 --> 00:22:12.974
I think Cursor had nineteen percent.
00:22:13.335 --> 00:22:14.375
It was still growing because I think,
00:22:14.395 --> 00:22:15.757
like you said, the market was massive.
00:22:15.817 --> 00:22:18.278
But I have a funny feeling, though.
00:22:18.479 --> 00:22:20.099
I mean,
00:22:20.119 --> 00:22:21.560
I knew a lot of people that started
00:22:21.621 --> 00:22:24.423
on Cursor who then moved over to Claude
00:22:24.462 --> 00:22:26.105
Code and a lot of people that were
00:22:26.204 --> 00:22:26.964
using Cursor,
00:22:26.984 --> 00:22:28.686
but actually just using the Claude Code
00:22:28.727 --> 00:22:29.186
terminal,
00:22:29.207 --> 00:22:30.627
just using it as an IDE and then
00:22:30.647 --> 00:22:31.888
stopped paying for Cursor because they
00:22:31.909 --> 00:22:33.210
weren't really using their model anymore.
00:22:35.750 --> 00:22:37.070
And I imagine that trajectory will
00:22:37.090 --> 00:22:38.011
probably increase.
00:22:38.071 --> 00:22:40.811
And even like today,
00:22:40.832 --> 00:22:42.232
a lot of the customers we work with,
00:22:43.053 --> 00:22:44.633
all I've ever heard is call code more
00:22:44.653 --> 00:22:45.374
than cursor.
00:22:45.433 --> 00:22:46.693
I don't hear anybody saying, hey,
00:22:46.733 --> 00:22:48.493
how can I roll out cursor safely?
00:22:50.615 --> 00:22:51.634
And what's the process for that?
00:22:51.674 --> 00:22:54.875
It seems to be because Anthropix more of
00:22:54.955 --> 00:22:55.796
general model,
00:22:56.336 --> 00:22:58.537
not just a development dev tooling
00:22:58.596 --> 00:22:59.096
centric.
00:22:59.778 --> 00:23:00.077
Obviously,
00:23:00.117 --> 00:23:02.557
it makes more sense to have an agnostic
00:23:02.597 --> 00:23:04.479
approach to everything.
00:23:05.492 --> 00:23:08.413
development for your business um you know
00:23:08.433 --> 00:23:11.074
for your genetic services for you know you
00:23:11.114 --> 00:23:14.634
don't so i i don't know it felt
00:23:14.654 --> 00:23:16.816
like an odd thing unless he's got a
00:23:16.836 --> 00:23:20.675
desires to to make the same his model
00:23:20.715 --> 00:23:22.136
which is obviously the thing is it's
00:23:22.196 --> 00:23:24.997
obvious to use that skill to get you
00:23:25.017 --> 00:23:27.897
know um xai's model to be also a
00:23:28.018 --> 00:23:30.278
universal model for everything as well and
00:23:30.317 --> 00:23:32.019
help train it on lots of things i
00:23:32.078 --> 00:23:32.439
presume
00:23:33.294 --> 00:23:34.653
But John,
00:23:35.153 --> 00:23:38.615
you're forgetting some things that we will
00:23:38.654 --> 00:23:41.915
never have the information to decide to
00:23:41.955 --> 00:23:43.036
make our own opinion, right?
00:23:43.477 --> 00:23:46.356
Yes, Claude is all the rage today,
00:23:47.176 --> 00:23:48.238
but is it going to be all the
00:23:48.278 --> 00:23:48.897
rage tomorrow?
00:23:49.258 --> 00:23:49.738
We don't know.
00:23:50.337 --> 00:23:51.097
Here we go.
00:23:51.198 --> 00:23:52.878
Anti-anthropic every time.
00:23:54.775 --> 00:23:56.115
No, I'm just talking about this, right?
00:23:56.836 --> 00:23:59.140
Which is why everybody needs to use
00:23:59.339 --> 00:23:59.920
OpenAI.
00:24:00.340 --> 00:24:02.423
God's sake, everyone.
00:24:04.105 --> 00:24:06.768
No, but remember, John,
00:24:06.847 --> 00:24:07.848
last week what happened?
00:24:07.909 --> 00:24:09.490
The new model came out, Fable.
00:24:10.010 --> 00:24:11.031
Everybody lost their mind.
00:24:11.231
--> 00:24:11.873
I lost my mind.
00:24:11.893 --> 00:24:13.994
I was like, wow, this thing is amazing.
00:24:14.715 --> 00:24:16.337
Next morning I woke up, bam.
00:24:17.228 --> 00:24:18.709
models not available in your country
00:24:18.749 --> 00:24:20.148
because you're not an american citizen
00:24:20.548 --> 00:24:23.150
right so why was that banned i don't
00:24:23.170 --> 00:24:25.131
know john but also at the same time
00:24:25.310 --> 00:24:28.112
it doesn't the tokens for fable were
00:24:28.152 --> 00:24:29.991
really expensive that model is super
00:24:30.051 --> 00:24:34.114
expensive if you come out as cursor
00:24:35.417 --> 00:24:38.599
feat featuring xai and your token
00:24:38.641 --> 00:24:40.221
consumption because token consumption is a
00:24:40.261 --> 00:24:41.863
problem now right things are getting
00:24:41.883 --> 00:24:44.605
really expensive and if you can somehow
00:24:45.405 --> 00:24:46.987
say you know what this model is going
00:24:47.007 --> 00:24:48.728
to cost maybe it's going to be even
00:24:48.748 --> 00:24:50.970
a bit cheaper than what anthropic model is
00:24:51.490 --> 00:24:53.632
perhaps you can get some people in your
00:24:54.512 --> 00:24:55.653
you can get some of the market share
00:24:56.730 --> 00:24:58.211
So all I'm saying, John,
00:24:58.711 --> 00:25:00.712
is that I'm not anti-any company.
00:25:00.732 --> 00:25:02.094
I'm not anti-anthropic.
00:25:02.674 --> 00:25:04.516
But what's happening today is what not
00:25:04.615 --> 00:25:06.117
necessarily is going to happen tomorrow.
00:25:06.577 --> 00:25:08.058
So I think it's a good gamble,
00:25:08.599 --> 00:25:09.200
in my opinion.
00:25:10.141 --> 00:25:10.961
I'm calling it a gamble,
00:25:10.980 --> 00:25:12.782
but it's a business decision because
00:25:13.624 --> 00:25:15.805
SpaceX valuation is two trillion.
00:25:15.924 --> 00:25:16.786
And as you said, what,
00:25:17.125 --> 00:25:19.327
sixty billions or roughly three percent of
00:25:19.407 --> 00:25:20.368
the equity slice.
00:25:23.250 --> 00:25:24.932
imagine if you spend sixty billion you end
00:25:24.951 --> 00:25:28.734
up getting two hundred billion why not why
00:25:28.755 --> 00:25:30.395
not because you know you've got the space
00:25:30.796 --> 00:25:32.917
side covered you've got you already have
00:25:33.097 --> 00:25:35.739
XAI which is like the X and the
00:25:35.759 --> 00:25:38.381
grok perhaps increasing that but you know
00:25:38.401 --> 00:25:39.862
that's it'll be interesting to see what
00:25:39.882 --> 00:25:42.443
happens in future do you think you will
00:25:42.463 --> 00:25:45.766
manage to achieve building the data
00:25:45.806 --> 00:25:47.747
centers out in space and then running
00:25:47.767 --> 00:25:50.249
those models out in space and then using
00:25:50.348 --> 00:25:51.029
Starlink to
00:25:51.750 --> 00:25:53.690
then obviously stream back down and do you
00:25:53.769 --> 00:25:58.391
reckon that's
the yeah so so even in
00:25:58.411 --> 00:26:00.251
their own ipo papers they said there's a
00:26:00.291 --> 00:26:02.132
lot of challenges around running this
00:26:02.152 --> 00:26:03.771
right so when the papers came out there's
00:26:03.791 --> 00:26:05.653
a lot a lot of challenges but he's
00:26:05.752 --> 00:26:07.272
it was a lot of challenges with spacex
00:26:07.913 --> 00:26:11.013
really to be fair like everyone yeah you
00:26:11.173 --> 00:26:12.433
know it's like to be able to do
00:26:12.473 --> 00:26:14.034
that and to land back down and be
00:26:14.094 --> 00:26:16.173
so precise and you know all those
00:26:16.193 --> 00:26:19.015
challenges and make it that cheap and um
00:26:20.228 --> 00:26:23.089
Yeah, I think technically, no, it's true.
00:26:23.170 --> 00:26:23.470
It's true.
00:26:23.509 --> 00:26:24.971
But the problem is, well,
00:26:24.990 --> 00:26:25.571
it's not a problem,
00:26:25.592 --> 00:26:27.093
but there's a slight difference because,
00:26:27.432 --> 00:26:29.434
yes, making it reusable was amazing.
00:26:29.795 --> 00:26:31.236
But I think there's like radiation is a
00:26:31.276 --> 00:26:32.517
problem as well up in space.
00:26:32.797 --> 00:26:35.818
But I think the reason why SpaceX was
00:26:35.858 --> 00:26:38.461
successful is because SpaceX needed that
00:26:38.500 --> 00:26:40.903
technology to send the satellites out
00:26:40.942 --> 00:26:42.403
there to land people on the moon, right?
00:26:42.903 --> 00:26:44.244
But data centers,
00:26:44.545 --> 00:26:45.805
there's still a lot of land that needs
00:26:45.825 --> 00:26:46.305
to be explored.
00:26:46.325 --> 00:26:46.505
You know,
00:26:46.545 --> 00:26:48.185
we talked about you can run your data
00:26:48.226 --> 00:26:49.046
centers in the ocean.
00:26:49.086 --> 00:26:50.646
And I know, John,
00:26:50.666 --> 00:26:52.007
that you want to swim next to a
00:26:52.047 --> 00:26:52.948
data center in a sea.
00:26:52.968 --> 00:26:54.888
I know that's one of your bucket list
00:26:54.989 --> 00:26:55.269
items.
00:26:55.828 --> 00:26:57.589
So I don't know if they will focus
00:26:57.650 --> 00:27:01.111
on running data centers in space because
00:27:01.351 --> 00:27:05.712
you want to run some stuff on Earth
00:27:05.732 --> 00:27:06.012
first.
00:27:06.553 --> 00:27:08.473
When you run out of space, then perhaps.
00:27:09.433 --> 00:27:09.753
Yeah, I mean,
00:27:09.773 --> 00:27:10.574
that's the way to look at it,
00:27:10.594 --> 00:27:10.753
isn't it?
00:27:10.874 --> 00:27:11.775
Like, let's really...
00:27:12.932 --> 00:27:16.395
the best way to align everybody to
00:27:16.435 --> 00:27:18.617
building data centers in space or just
00:27:18.657 --> 00:27:21.801
space in general is to ruin us first.
00:27:22.742 --> 00:27:24.104
If we can get everybody
00:27:25.365 --> 00:27:27.025
aligned on that if we can ruin it
00:27:27.065 --> 00:27:28.925
first then we all want to go to
00:27:28.986 --> 00:27:31.246
space we're like all on board we're all
00:27:31.286 --> 00:27:32.766
aligned on that vision we're like yeah i
00:27:32.786 --> 00:27:34.626
mean who wants to live here now it's
00:27:34.686 --> 00:27:37.247
totally destroyed i at least we then share
00:27:37.267 --> 00:27:38.688
the common vision of like how do we
00:27:38.708 --> 00:27:41.647
support spacex and getting us to space now
00:27:41.688 --> 00:27:44.828
that we've uh you know destroyed earth i
00:27:44.868 --> 00:27:46.969
suppose it becomes a common maybe we can
00:27:47.009 --> 00:27:48.949
get as a hotel there as well so
00:27:48.969 --> 00:27:50.450
you're like you're all begging elon like
00:27:50.509 --> 00:27:51.650
get us off this planet
00:27:52.673 --> 00:27:54.315
Get me a seat on SpaceX.
00:27:54.335 --> 00:27:55.194
Get me off this planet.
00:27:55.556 --> 00:27:57.076
We'll get you into your flying Tesla.
00:27:57.737 --> 00:28:02.999
You're not wrong, John.
00:28:03.200 --> 00:28:05.181
Amazon came out last week.
00:28:05.601 --> 00:28:06.240
Well, actually,
00:28:06.260 --> 00:28:07.622
they came out last week and they said
00:28:07.721 --> 00:28:10.182
their data centers use about two point
00:28:10.222 --> 00:28:13.365
five billion gallon liter of water.
00:28:13.765 --> 00:28:15.105
Billion gallons liter of water.
00:28:15.685 --> 00:28:16.567
That's two point five.
00:28:16.807 --> 00:28:18.708
And same with the same with I think
00:28:18.748 --> 00:28:21.410
Google had similar numbers and I mean,
00:28:21.509 --> 00:28:22.790
a little bit less or whatnot.
00:28:23.191 --> 00:28:23.851
So you're right.
00:28:24.511 --> 00:28:26.913
We are ruining the earth.
00:28:27.334 --> 00:28:28.555
I think they are saying that they want
00:28:28.575 --> 00:28:31.537
to become by twenty thirty water positive.
00:28:31.557 --> 00:28:33.018
So as in like they don't waste the
00:28:33.038 --> 00:28:34.598
water more than they use.
00:28:35.019 --> 00:28:37.421
But yeah, this is becoming the efficiency.
00:28:37.780 --> 00:28:39.061
This is how we're already ruining it,
00:28:39.122 --> 00:28:39.301
right?
00:28:39.501 --> 00:28:40.383
We're already ruining it.
00:28:40.403 --> 00:28:41.762
Like body positive.
00:28:42.624 --> 00:28:43.805
You become water positive.
00:28:44.525 --> 00:28:45.566
Water positive, John.
00:28:45.826 --> 00:28:48.428
yeah and what about this uh what about
00:28:48.448 --> 00:28:51.790
this microsoft um capacity issue where
00:28:51.830 --> 00:28:53.271
it's actually having to run github in
00:28:53.332 --> 00:28:54.573
amazon what what have you got to say
00:28:54.593 --> 00:28:57.576
about that so john we're talking about
00:28:57.655 --> 00:29:00.057
ruining the earth right uh we need
00:29:00.357 --> 00:29:02.579
capacity because we're making so many data
00:29:02.640 --> 00:29:04.682
centers we're not finished ruining the
00:29:04.741 --> 00:29:06.323
earth itself so we
started moving into
00:29:06.363 --> 00:29:08.785
water with data centers in water hopefully
00:29:08.825 --> 00:29:11.366
coming soon and then maybe space as well
00:29:11.406 --> 00:29:12.208
but here's the problem
00:29:13.088 --> 00:29:13.829
So GitHub,
00:29:14.269 --> 00:29:16.832
our favorite source control solution,
00:29:17.031 --> 00:29:18.752
one of our favorite source control
00:29:18.792 --> 00:29:19.153
solution.
00:29:19.472 --> 00:29:20.273
In twenty eighteen,
00:29:20.314 --> 00:29:22.615
Microsoft bought it and they said by
00:29:22.755 --> 00:29:26.377
twenty twenty seven will fully migrate it
00:29:26.438 --> 00:29:27.659
to Azure.
00:29:27.679 --> 00:29:28.239
Right.
00:29:28.278 --> 00:29:28.940
Because, you know,
00:29:28.960 --> 00:29:29.880
they were running some stuff in AWS
00:29:29.900 --> 00:29:30.039
before.
00:29:30.059 --> 00:29:30.160
Yeah.
00:29:30.180 --> 00:29:30.259
Yeah.
00:29:30.279 --> 00:29:30.881
They still got it.
00:29:31.580 --> 00:29:33.823
Yeah, they've got it.
00:29:33.843 --> 00:29:33.923
Yeah.
00:29:37.204 --> 00:29:37.905
So here is.
00:29:38.891 --> 00:29:40.913
Everything's related, John.
00:29:41.553 --> 00:29:42.693
So they were running this stuff.
00:29:42.953 --> 00:29:43.555
They were fine.
00:29:43.694 --> 00:29:47.337
They were on track to move on from
00:29:47.438 --> 00:29:48.439
AWS to Azure.
00:29:50.000 --> 00:29:52.102
But your favorite company,
00:29:52.162 --> 00:29:53.643
Anthropic and OpenAI,
00:29:53.863 --> 00:29:54.864
came into the picture.
00:29:55.845 --> 00:29:57.506
And they told developers, oh,
00:29:57.747 --> 00:29:58.727
in twenty twenty five,
00:29:59.248 --> 00:30:04.291
you committed one billion commits of code.
00:30:04.332 --> 00:30:04.791
That's cool,
00:30:05.133 --> 00:30:07.355
but let's have a bit more slop.
00:30:08.060 --> 00:30:08.821
Right.
00:30:08.882 --> 00:30:09.903
Not just a bit more,
00:30:10.223 --> 00:30:12.646
fourteen times more slop in our
00:30:12.666 --> 00:30:14.088
repositories than we had before.
00:30:14.549 --> 00:30:16.712
I feel like you're aiming this at me
00:30:16.873 --> 00:30:18.134
and my PRs or something.
00:30:18.214 --> 00:30:19.135
It feels really personal.
00:30:19.155 --> 00:30:22.220
The way you directed that felt really
00:30:23.040 --> 00:30:24.362
Like an attack on me, actually,
00:30:24.382 --> 00:30:25.021
a little bit there.
00:30:25.502 --> 00:30:28.045
Have you been reviewing my code or
00:30:28.065 --> 00:30:28.484
something?
00:30:28.525 --> 00:30:30.865
I've been reviewing your code, right?
00:30:31.207 --> 00:30:32.708
I've been reviewing your code.
00:30:33.387 --> 00:30:36.089
So the code commits within one year have
00:30:36.150 --> 00:30:38.692
gone from one billion to fourteen billion
00:30:38.751 --> 00:30:39.152
commits.
00:30:41.272 --> 00:30:44.953
So the scale is massively increased.
00:30:45.453 --> 00:30:46.733
And there were a lot of issues.
00:30:46.855 --> 00:30:49.654
You might be aware of GitHub going down
00:30:49.734 --> 00:30:50.775
because of the scale.
00:30:51.175 --> 00:30:52.935
It was kind of not predicted that it's
00:30:52.955 --> 00:30:54.715
going to be at that scale.
00:30:54.776 --> 00:30:57.836
So basically what's happening is at this
00:30:57.876 --> 00:30:58.297
scale,
00:30:58.856 --> 00:31:01.317
Microsoft do need extra capacity to run
00:31:01.337 --> 00:31:01.538
this.
00:31:03.481 --> 00:31:04.702
And hence why they're like, okay,
00:31:04.722 --> 00:31:05.083
you know what?
00:31:05.282 --> 00:31:06.263
We do need to use,
00:31:06.545 --> 00:31:08.145
maybe stick with a bit of AWS because
00:31:08.185 --> 00:31:09.366
we're getting all these outages.
00:31:09.406 --> 00:31:10.847
We're running out of capacity.
00:31:11.367 --> 00:31:13.069
So this is a bit of a problem
00:31:13.130 --> 00:31:15.731
that this is being produced because of all
00:31:15.791 --> 00:31:19.575
these bots, commits that are coming,
00:31:19.615 --> 00:31:22.237
not just developer-assisted commits,
00:31:22.257 --> 00:31:23.999
but also agents committing code
00:31:24.019 --> 00:31:24.479
themselves.
00:31:25.992 --> 00:31:28.655
So this is a bit of a problem
00:31:28.695 --> 00:31:31.618
for Microsoft because GitHub is losing a
00:31:31.638 --> 00:31:32.539
bit of market share.
00:31:34.162 --> 00:31:37.526
And the point here is that this is
00:31:37.566 --> 00:31:38.507
an architecture lesson,
00:31:38.527 --> 00:31:42.171
that even your own cloud isn't infinitely
00:31:42.310 --> 00:31:43.332
elastic on demand.
00:31:44.153 --> 00:31:46.256
And multi-cloud is sometimes, sometimes,
00:31:47.555 --> 00:31:49.596
reasonable solution and that's what that's
00:31:49.615 --> 00:31:51.997
what microsoft is saying themselves oh
00:31:52.096 --> 00:31:53.676
maybe we need to be multi-cloud i mean
00:31:53.717 --> 00:31:56.157
but that scale is it's completely bonkers
00:31:56.198 --> 00:31:58.358
right eighteen fourteen billion commits
00:31:58.419 --> 00:32:00.000
and a bunch of other things are happening
00:32:00.539 --> 00:32:02.181
so this is uh so that's what's happening
00:32:02.221 --> 00:32:04.161
john all these things are related with
00:32:04.201 --> 00:32:07.522
each other and perhaps you're right maybe
00:32:07.542 --> 00:32:08.323
we do need those
00:32:09.022 --> 00:32:12.246
satellite data centers they're flying
00:32:12.266 --> 00:32:15.567
around do you reckon github because of the
00:32:15.607 --> 00:32:18.549
cost to them on hosting now will put
00:32:18.589 --> 00:32:20.510
the price up of github surely they must
00:32:20.611 --> 00:32:22.512
must do because the amount of like the
00:32:22.553 --> 00:32:25.375
cost to like what's the margin on all
00:32:25.414 --> 00:32:26.595
of that now they're going especially if
00:32:26.615 --> 00:32:27.895
they're bursting into amazon they've got
00:32:27.915 --> 00:32:31.098
to pay amazon for all that compute it'd
00:32:31.118 --> 00:32:32.358
be interesting to see if they put the
00:32:32.400 --> 00:32:32.880
prices up
00:32:33.935 --> 00:32:36.398
Prices up for GitHub Enterprise.
00:32:36.459 --> 00:32:37.740
Prices are for Copilot.
00:32:37.759 --> 00:32:38.000
Yeah,
00:32:38.059 --> 00:32:39.040
I think there is going to be an
00:32:39.080 --> 00:32:40.423
adjustment, John, for sure.
00:32:40.442 --> 00:32:41.463
Is it going to go up?
00:32:42.243 --> 00:32:43.005
Is it going to go down?
00:32:43.385 --> 00:32:43.865
No idea.
00:32:43.925 --> 00:32:46.648
But inflation is also kicking in as
00:32:46.729 --> 00:32:47.009
always.
00:32:47.028 --> 00:32:49.372
So the price will only go up,
00:32:49.451 --> 00:32:49.731
I guess.
00:32:50.133 --> 00:32:50.452
But yeah,
00:32:50.492 --> 00:32:51.453
this is a bit of an issue.
00:32:52.134 --> 00:32:56.179
But the thing is that this is the
00:32:56.219 --> 00:32:56.939
reflection point.
00:32:57.901 --> 00:32:58.181
This,
00:32:58.641 --> 00:33:02.582
we've seen AI infrastructure on demand is
00:33:02.682 --> 00:33:03.903
really increasing a lot.
00:33:04.403 --> 00:33:06.222
But we do need a bunch of that
00:33:06.262 --> 00:33:06.584
scale.
00:33:06.663 --> 00:33:08.084
It's not just for training the models,
00:33:08.124 --> 00:33:09.744
John, right?
00:33:09.765 --> 00:33:11.025
It's not just for those tools that are
00:33:11.045 --> 00:33:12.204
going to give you all that slop.
00:33:12.404 --> 00:33:15.385
But it's also for when you produce all
00:33:15.405 --> 00:33:16.727
that code, where you host it,
00:33:16.747 --> 00:33:17.366
where it runs.
00:33:17.846 --> 00:33:19.547
So the demand is only going to go
00:33:19.567 --> 00:33:19.928
up, John.
00:33:20.307 --> 00:33:22.008
I think we should change and go into
00:33:22.067 --> 00:33:23.368
data center technology, John.
00:33:23.729 --> 00:33:24.588
I think that's what we need to do.
00:33:25.398 --> 00:33:26.559
We need to get into chips.
00:33:27.361 --> 00:33:27.641
Chips?
00:33:28.020 --> 00:33:29.801
Not data chips, obviously, you know,
00:33:29.842 --> 00:33:31.163
like GPU chips.
00:33:32.304 --> 00:33:34.905
We need to get into chips and data
00:33:34.925 --> 00:33:35.286
centers.
00:33:35.306 --> 00:33:35.767
You're right.
00:33:36.086 --> 00:33:38.409
That's where we need to do.
00:33:38.429 --> 00:33:38.469
Oh,
00:33:38.509 --> 00:33:42.231
we can get into the other technology that,
00:33:43.732 --> 00:33:44.012
who was it,
00:33:44.032 --> 00:33:46.434
NVIDIA had where you can strap it to
00:33:46.454 --> 00:33:47.434
your house, basically.
00:33:47.536 --> 00:33:50.518
Just like a mini data center strapped to
00:33:50.557 --> 00:33:51.298
the side of your house.
00:33:51.317 --> 00:33:51.499
That's...
00:33:52.482 --> 00:33:53.784
That's the way forward.
00:33:53.903 --> 00:33:54.104
Yeah.
00:33:54.565 --> 00:33:57.227
Like a solar paneled data center
on the
00:33:57.267 --> 00:33:59.890
roof of your house or something.
00:34:00.911 --> 00:34:01.131
Maybe.
00:34:01.530 --> 00:34:02.271
Everything is compute.
00:34:02.571 --> 00:34:03.853
But what about this Commodore?
00:34:04.794 --> 00:34:05.634
Commodore is back.
00:34:05.674 --> 00:34:06.736
Not the Commodore sixty four.
00:34:08.036 --> 00:34:10.659
This is the Commodore flip phone.
00:34:11.501 --> 00:34:12.061
Flip phone.
00:34:12.320 --> 00:34:13.041
That has come back.
00:34:13.101 --> 00:34:13.342
Yeah.
00:34:13.621 --> 00:34:14.702
And running Android.
00:34:15.905 --> 00:34:17.405
Are you going to buy one?
00:34:18.123 --> 00:34:18.782
I'm going to buy one.
00:34:18.822 --> 00:34:19.163
Why not?
00:34:19.264 --> 00:34:21.364
It's a Linux powered flip phone.
00:34:22.405 --> 00:34:23.985
It's going to cost around five hundred
00:34:24.025 --> 00:34:24.365
dollars.
00:34:25.126 --> 00:34:26.666
So it prevents.
00:34:28.067 --> 00:34:29.608
This is the this is the kicker.
00:34:29.688 --> 00:34:31.849
It prevents social media app installation.
00:34:32.289 --> 00:34:32.568
John,
00:34:32.909 --> 00:34:33.849
I know it's not going to work for
00:34:33.869 --> 00:34:33.969
you.
00:34:34.376 --> 00:34:35.416
Because you won't be able to install
00:34:35.456 --> 00:34:35.657
TikTok.
00:34:35.757 --> 00:34:36.057
I'm out.
00:34:37.217 --> 00:34:39.059
You're out.
00:34:39.079 --> 00:34:39.458
Yeah, I'm out.
00:34:39.478 --> 00:34:40.500
But this sounds really good.
00:34:40.840 --> 00:34:42.820
It's got a forty-eight megapixel camera.
00:34:42.960 --> 00:34:45.121
We've now become tech reviewers now, John,
00:34:45.141 --> 00:34:45.282
right?
00:34:45.302 --> 00:34:46.882
So that's what we're talking about.
00:34:47.222 --> 00:34:50.304
But the point here is that with UK
00:34:50.344 --> 00:34:52.264
government deciding that we need to ban
00:34:52.284 --> 00:34:53.664
social media for under-sixteens,
00:34:53.865 --> 00:34:55.846
I don't know what your take on that
00:34:55.925 --> 00:34:56.065
is.
00:34:56.407 --> 00:34:57.007
I agree with it.
00:34:58.688 --> 00:34:59.668
Plus, I'm not under-sixteen,
00:34:59.688 --> 00:35:00.367
so it's okay for me.
00:35:00.387 --> 00:35:01.048
Do you agree with it?
00:35:01.068 --> 00:35:02.188
I'm not surprised.
00:35:02.208 --> 00:35:02.989
I agree with it.
00:35:03.090 --> 00:35:04.050
I agree with it.
00:35:04.940 --> 00:35:09.344
I agree with the fact that generations
00:35:09.364 --> 00:35:12.025
have grown up without social media.
00:35:12.045 --> 00:35:14.206
And I think that was probably a positive
00:35:14.246 --> 00:35:14.487
thing.
00:35:15.527 --> 00:35:18.329
But I don't necessarily agree with the cop
00:35:18.449 --> 00:35:21.751
out of not forcing companies to legislate
00:35:21.971 --> 00:35:23.932
against protecting people on the platform.
00:35:24.072 --> 00:35:25.914
It feels like the people penalized.
00:35:26.822 --> 00:35:30.385
for the issues is now the children rather
00:35:30.405 --> 00:35:33.268
than making the adjustments in the right
00:35:33.309 --> 00:35:35.331
places to those platforms to begin with.
00:35:35.411 --> 00:35:36.351
It feels like they kind of got,
00:35:36.371 --> 00:35:37.592
I don't know,
00:35:38.273 --> 00:35:39.235
it feels like there's going to be
00:35:39.255 --> 00:35:40.655
bypassing and they've kind of got,
00:35:40.675 --> 00:35:41.836
the people are going to get around it
00:35:42.358 --> 00:35:42.657
and
00:35:43.594 --> 00:35:45.456
The cultures have shifted, right?
00:35:45.476 --> 00:35:46.536
We're just talking about AI.
00:35:46.635 --> 00:35:48.657
We're talking about bloody rockets.
00:35:49.496 --> 00:35:51.157
We're talking about data centers in space.
00:35:51.438 --> 00:35:52.498
The world isn't quite the same.
00:35:52.539 --> 00:35:54.239
People are swimming next to data centers.
00:35:54.719 --> 00:35:58.701
So everything's kind of altering.
00:36:00.161 --> 00:36:01.621
Social media has been a big part of
00:36:01.641 --> 00:36:02.882
that for good or for bad.
00:36:02.963 --> 00:36:04.503
Probably more bad than good, I would say.
00:36:04.523 --> 00:36:05.684
But anyway, that's just my opinion.
00:36:06.840 --> 00:36:08.181
They could have adjusted it, though,
00:36:10.161 --> 00:36:13.585
and made it more safe somehow and got
00:36:14.365 --> 00:36:15.907
them to change it and legislate,
00:36:16.427 --> 00:36:19.769
not be like, well, I suppose, you know.
00:36:19.849 --> 00:36:21.050
It would be like going on holiday.
00:36:22.231 --> 00:36:24.052
Somebody dies from alcohol poisoning.
00:36:24.072 --> 00:36:24.932
They're like, you know what?
00:36:24.972 --> 00:36:26.273
Just stop kids going on holiday.
00:36:26.974 --> 00:36:28.775
Let's just stop all kids going on holiday.
00:36:30.117 --> 00:36:32.559
Nobody goes abroad until you're eighteen.
00:36:33.400 --> 00:36:33.880
We can't have...
00:36:33.900 --> 00:36:34.601
Do you know what I mean?
00:36:34.681 --> 00:36:35.981
It's such an extreme...
00:36:36.724 --> 00:36:42.786
view to the problem.
00:36:43.927 --> 00:36:44.387
Fair enough.
00:36:44.586 --> 00:36:47.288
That's because the government legislation,
00:36:47.608 --> 00:36:50.268
they've been asking the companies to add
00:36:50.329 --> 00:36:51.469
in more guardrails.
00:36:51.489 --> 00:36:52.710
They do add guardrails like these
00:36:52.750 --> 00:36:53.130
companies,
00:36:53.610 --> 00:36:55.791
Instagram and Snapchat or whatever,
00:36:56.132 --> 00:36:57.632
that you need to be a certain age
00:36:57.652 --> 00:36:58.373
and that before you...
00:36:59.172 --> 00:36:59.913
before you can log in,
00:36:59.952 --> 00:37:01.733
but it's kind of hard to police that.
00:37:02.375 --> 00:37:04.335
How would you stop anybody not like,
00:37:04.456 --> 00:37:05.476
I can pick your phone, John,
00:37:05.817 --> 00:37:07.396
if I know your
passcode and I can
00:37:07.436 --> 00:37:09.978
pretend it's you or log in using your
00:37:09.998 --> 00:37:10.478
credentials.
00:37:10.518 --> 00:37:11.840
People work your way around.
00:37:12.440 --> 00:37:12.679
I think
00:37:14.440 --> 00:37:15.862
You can argue both ways.
00:37:16.483 --> 00:37:17.483
I agree with what you're saying.
00:37:17.862 --> 00:37:20.164
But I think a blanket ban is probably
00:37:20.184 --> 00:37:22.266
better because there's so much rubbish on
00:37:22.306 --> 00:37:24.027
social media nowadays and the pressure
00:37:24.047 --> 00:37:26.809
that it puts on the teenagers onto doing
00:37:27.009 --> 00:37:27.909
specific things,
00:37:27.949 --> 00:37:29.210
following specific trends.
00:37:30.411 --> 00:37:31.632
Australia did it, right?
00:37:31.672 --> 00:37:33.132
They started that a few months ago.
00:37:33.793 --> 00:37:34.713
I think it'll be good to see what
00:37:34.733 --> 00:37:34.974
happens.
00:37:35.014 --> 00:37:35.574
But I agree.
00:37:35.594 --> 00:37:35.835
That's why...
00:37:37.458 --> 00:37:41.760
I have respect for making a decision and
00:37:41.840 --> 00:37:43.643
actually coming down on a line against
00:37:43.682 --> 00:37:50.027
something that is maybe not healthy for
00:37:50.186 --> 00:37:53.070
kids and just generally their state of
00:37:53.130 --> 00:37:53.349
mind.
00:37:53.409 --> 00:37:54.690
I kind of agree with that.
00:37:57.630 --> 00:38:00.014
It just feels like the wrong answer to
00:38:00.034 --> 00:38:00.775
the problem to me.
00:38:00.815 --> 00:38:01.077
But anyway.
00:38:01.137 --> 00:38:03.119
John, John, John, let me, let me,
00:38:03.219 --> 00:38:04.541
let me stop, finish with this, right?
00:38:05.382 --> 00:38:08.728
So UK government mandates that everybody
00:38:08.789 --> 00:38:10.070
under sixteen has to go to school.
00:38:11.420 --> 00:38:11.699
Agreed?
00:38:12.340 --> 00:38:12.960
That's a mandate.
00:38:13.161 --> 00:38:15.400
Everybody under under-sixteen has to have
00:38:15.420 --> 00:38:16.081
an education.
00:38:16.981 --> 00:38:19.063
You could be homeschooled though,
00:38:19.083 --> 00:38:19.322
can't you?
00:38:19.382 --> 00:38:20.882
You can be homeschooled, correct.
00:38:21.202 --> 00:38:22.344
But you have to be schooled.
00:38:23.063 --> 00:38:25.605
But that homeschooling is by exception.
00:38:25.644 --> 00:38:27.326
So are we saying that when the government
00:38:27.365 --> 00:38:28.846
puts in a rule that says everybody has
00:38:28.865 --> 00:38:31.746
to go to school, nah, don't do it.
00:38:32.126 --> 00:38:32.947
Let parents decide.
00:38:33.807 --> 00:38:34.748
Figure out how to do it.
00:38:35.068 --> 00:38:36.208
But here's the thing though.
00:38:36.849 --> 00:38:37.648
If a child says,
00:38:37.889 --> 00:38:39.090
I don't really feel like going to school.
00:38:39.309 --> 00:38:39.490
Well,
00:38:39.510 --> 00:38:40.530
you have to go to school because the
00:38:40.550 -->
00:38:41.891
government says you have to go to school.
00:38:42.351 --> 00:38:42.692
Otherwise,
00:38:42.711 --> 00:38:43.472
they're going to put you in jail.
00:38:44.092 --> 00:38:45.514
So now it's become a lot easier for
00:38:45.554 --> 00:38:45.974
parents.
00:38:46.153 --> 00:38:46.795
If they say, oh,
00:38:46.815 --> 00:38:47.954
can I have a Snapchat account?
00:38:48.255 --> 00:38:49.576
Well, actually, you can't.
00:38:49.596 --> 00:38:51.197
Did you think people weren't going to
00:38:51.277 --> 00:38:52.958
school before that mandate?
00:38:53.378 --> 00:38:54.219
Is that what the problem was?
00:38:54.838 --> 00:38:55.139
Everybody,
00:38:55.159 --> 00:38:56.940
because you'd be walking around.
00:38:57.179 --> 00:38:57.960
No one's at school.
00:38:58.181 --> 00:38:58.800
Do you remember that?
00:38:58.840 --> 00:38:59.961
When you used to walk around?
00:38:59.981 --> 00:39:00.061
Yeah.
00:39:01.402 --> 00:39:02.463
All the schools were empty.
00:39:02.483 --> 00:39:03.422
Do you remember those days?
00:39:03.663 --> 00:39:05.384
All the schools were empty.
00:39:05.764 --> 00:39:06.884
All the kids were just out on the
00:39:06.903 --> 00:39:07.264
streets.
00:39:07.304 --> 00:39:08.465
And then the government came along and
00:39:08.485 --> 00:39:10.164
were like, you need to be at school.
00:39:10.304 --> 00:39:11.846
And then everyone started to go to school.
00:39:11.945 --> 00:39:14.367
I mean, that's not what happened.
00:39:14.626 --> 00:39:15.887
They didn't need to mandate it.
00:39:15.907 --> 00:39:17.007
There wasn't really a problem,
00:39:17.088 --> 00:39:17.648
I don't think.
00:39:17.688 --> 00:39:18.907
It was like, what was it for?
00:39:18.947 --> 00:39:20.188
Like, whatever percentage.
00:39:20.208 --> 00:39:21.429
It seems a bit like, yeah.
00:39:21.648 --> 00:39:24.269
I mean, bizarre to me,
00:39:24.289 --> 00:39:25.731
even mandating going to school.
00:39:25.911 --> 00:39:28.592
It seems a little bit silly because...
00:39:29.545 --> 00:39:30.144
Yeah,
00:39:30.164 --> 00:39:32.766
I don't think that you're a homeschool
00:39:32.786 --> 00:39:33.768
kind of guy.
00:39:33.827 --> 00:39:35.809
I'm a homeschool kind of person.
00:39:35.949 --> 00:39:36.929
I'm a social media.
00:39:37.610 --> 00:39:39.771
I believe that people should upskill on
00:39:39.871 --> 00:39:40.652
social media.
00:39:41.152 --> 00:39:43.893
The best way to get educated is to
00:39:43.934 --> 00:39:47.356
just get TikTok and to get Instagram.
00:39:47.876 --> 00:39:50.518
And you just scroll and you learn through
00:39:50.539 --> 00:39:51.039
scrolling.
00:39:51.759 --> 00:39:52.679
And the more you can scroll,
00:39:52.699 --> 00:39:53.601
the more you tend to learn.
00:39:54.708 --> 00:39:56.309
That's what's got me where I am today,
00:39:56.349 --> 00:39:56.748
Salman.
00:39:56.889 --> 00:39:58.509
Without social media and scrolling,
00:39:58.568 --> 00:40:02.510
I wouldn't be here where I am today.
00:40:02.530 --> 00:40:04.431
Welcome to the Diary of a CEO,
00:40:04.472 --> 00:40:08.193
where John tells how he got here.
00:40:09.981 --> 00:40:11.242
Yeah, I didn't go to school.
00:40:11.422 --> 00:40:12.802
It was all social media then.
00:40:12.983 --> 00:40:14.302
That's how I upskilled.
00:40:14.342 --> 00:40:15.824
But yeah, anyway, Commodore,
00:40:16.103 --> 00:40:17.724
you're going to get one.
00:40:18.543 --> 00:40:19.485
It looks really cool.
00:40:19.545 --> 00:40:20.505
It's a funky screen.
00:40:20.565 --> 00:40:21.764
It's like blue in color.
00:40:21.824 --> 00:40:23.525
We'll never hear from you ever again once
00:40:23.545 --> 00:40:24.346
you've got one, I think.
00:40:25.666 --> 00:40:26.947
But it doesn't have social media,
00:40:26.987 --> 00:40:28.067
so that's good.
00:40:28.586 --> 00:40:31.248
I can't be on TikTok.
00:40:31.387 --> 00:40:32.007
I don't know.
00:40:32.088 --> 00:40:32.367
Anyway.
00:40:34.405 --> 00:40:35.166
Until next time,
00:40:35.326 --> 00:40:36.507
I think the next one we're going to
00:40:36.867 --> 00:40:38.927
pick a topic and dig into it a
00:40:38.947 --> 00:40:39.608
little bit more.
00:40:39.648 --> 00:40:40.807
We could have been doing that anyway,
00:40:41.289 --> 00:40:42.628
a little bit on these news articles,
00:40:42.668 --> 00:40:45.969
but I think we'll probably go a bit
00:40:46.010 --> 00:40:49.251
deeper into something more specific going
00:40:49.311 --> 00:40:51.813
on in the industry on the next episode.
00:40:53.012 --> 00:40:53.293
But yeah,
00:40:53.693 --> 00:40:55.693
we shall speak to everybody next time.
00:40:57.235 --> 00:40:57.735
See you later.