TRANSKRIPT 🔗
1
00:00:00,000 --> 00:00:04,160
Welcome to Top 5 Tech, my weekly show about what to watch, listen to, read, download,
2
00:00:04,160 --> 00:00:05,559
and something deep every week.
3
00:00:05,559 --> 00:00:08,480
More of a top 10 this week because there's a bunch of great links that I wanted to point
4
00:00:08,480 --> 00:00:09,480
you to.
5
00:00:09,480 --> 00:00:13,640
The first one is an interview with Craig Federighi and YouTube channel Donald Decodes.
6
00:00:13,640 --> 00:00:18,120
He has some great questions about the iPhone Duo, swipe gestures, and Craig Federighi gets
7
00:00:18,120 --> 00:00:19,120
thankfully nerdy.
8
00:00:19,120 --> 00:00:22,559
Kind of explains the details about how they decide things like the backswipe and why it's
9
00:00:22,559 --> 00:00:25,199
not from both directions like on Android.
10
00:00:25,199 --> 00:00:26,199
Really fun interview.
11
00:00:26,199 --> 00:00:27,199
Not super long.
12
00:00:27,199 --> 00:00:28,199
It's about 11 minutes.
13
00:00:28,199 --> 00:00:29,199
So I'll link that.
14
00:00:30,000 --> 00:00:33,280
Then there's MUSE AI on primary technology and on last week on this show.
15
00:00:33,280 --> 00:00:36,259
And Joanna Stern's interview with Mark Zuckerberg is one of the best.
16
00:00:36,259 --> 00:00:40,280
She has hard-hitting questions both about the Ray-Bans, Metaglasses, and of course MUSE
17
00:00:40,280 --> 00:00:41,280
AI.
18
00:00:41,280 --> 00:00:42,920
So we'll link that interview down in the show notes as well.
19
00:00:42,920 --> 00:00:47,399
On what to listen to, I would listen with a big grain of salt, but Ezra Klein actually
20
00:00:47,399 --> 00:00:51,960
interviewed Bill Gates talking about AI and the warnings that Bill Gates have been giving
21
00:00:51,960 --> 00:00:52,960
very publicly.
22
00:00:52,960 --> 00:00:55,920
This is also a video interview if you want to watch it on the New York Times website.
23
00:00:55,920 --> 00:01:00,000
but I'm going to link the audio podcast in Apple Podcasts and Bill Gates.
24
00:01:00,000 --> 00:01:04,519
Bill Gates has some strong feelings. He thinks that while people are comparing the AI revolution
25
00:01:04,519 --> 00:01:09,319
right now to the industrial revolution, he thinks of it more as like an evolutionary
26
00:01:09,319 --> 00:01:14,200
leap. And he says many jobs are not just going to change in nature, but will be going away.
27
00:01:14,200 --> 00:01:18,799
I'm not going to say he's a doomsdayer, but he has some strong feelings about it. Now,
28
00:01:18,799 --> 00:01:22,400
we actually discussed on primary technology this week, how much we should take Bill Gates
29
00:01:22,400 --> 00:01:27,159
ideas and thoughts on AI, because he's not really in the world right now working on the
30
00:01:27,159 --> 00:01:30,079
frontier models, and Microsoft isn't really in the conversation.
31
00:01:30,000 --> 00:01:34,640
When it comes to all of that either, it's like open AI, Anthropic, Meta, Google.
Load full transcript
32
00:01:34,879 --> 00:01:38,519
So again, big grain of salt, but I did find it to be a fascinating interview.
33
00:01:38,519 --> 00:01:42,260
And Ezra Klein pushes back on a lot of the ideas that Bill Gates puts forward.
34
00:01:42,260 --> 00:01:43,480
So I thought it was a great interview.
35
00:01:43,560 --> 00:01:46,719
And on a second, what to listen to my co-host on primary technology, Jason
36
00:01:46,719 --> 00:01:49,960
Aiton interviewed Ron Johnson, which if you're not familiar with that name, he
37
00:01:49,960 --> 00:01:54,599
joined Apple and along with Steve Jobs helped build out the idea of the Apple
38
00:01:54,599 --> 00:01:57,519
store, the original concept, the genius bar.
39
00:01:57,760 --> 00:01:59,920
He came from Target and there's really fascinating
40
00:02:00,000 --> 00:02:04,000
stories about that and how Steve Jobs courted him to come to Apple.
41
00:02:04,000 --> 00:02:08,240
Several fun Steve Jobs stories in this episode but that's the episode of Creative Effort
42
00:02:08,240 --> 00:02:09,240
linked below.
43
00:02:09,240 --> 00:02:10,919
Now on what to read, I have two picks here.
44
00:02:10,919 --> 00:02:12,099
One's an actual read.
45
00:02:12,099 --> 00:02:15,860
One is you should read this website because a childhood friend of mine actually, Nick
46
00:02:15,860 --> 00:02:20,280
who I've known I think 35 years, he built Roadrunner.
47
00:02:20,280 --> 00:02:24,520
He vibe coded this and it is a better replacement for Nextdoor.
48
00:02:24,520 --> 00:02:29,500
Now this is only available in Catalina, Saddlebrook, Oro Valley, Oracle, San Manuel and the rest
49
00:02:30,000 --> 00:02:34,240
of the southern Tucson suburbs but it's a really cool website and he has ways where
50
00:02:34,240 --> 00:02:38,919
it's going to verify that these people actually live in this place, that it's private, newest
51
00:02:38,919 --> 00:02:39,919
posts first.
52
00:02:39,919 --> 00:02:44,379
Basically all the bloat and junk that makes Nextdoor not very fun to use, he built a way
53
00:02:44,379 --> 00:02:45,599
better option.
54
00:02:45,599 --> 00:02:48,840
Now it would be cool if he could expand this to many other neighborhoods, maybe even across
55
00:02:48,840 --> 00:02:49,840
the country.
56
00:02:49,840 --> 00:02:53,280
I'm not sure what that all would take but Roadrunner, especially if you live in that
57
00:02:53,280 --> 00:02:55,020
area, really fun tool.
58
00:02:55,020 --> 00:02:56,120
Great job Nick on building that.
59
00:02:56,120 --> 00:02:59,319
And I'll plug another one of Jason Aten's articles on what to read.
60
00:03:00,000 --> 00:03:04,960
Back in July, how the AI landscape right now is more like Jurassic Park than anything
61
00:03:04,960 --> 00:03:09,599
else and Jason put that into words over on Inc.com, how the AI industry has a Jurassic
62
00:03:09,599 --> 00:03:10,879
Park problem.
63
00:03:10,879 --> 00:03:11,879
Another great read.
64
00:03:11,879 --> 00:03:15,520
I
wanted to download, wanted to highlight two free apps, both completely free.
65
00:03:15,520 --> 00:03:17,960
The first one, Ensemble Music Player.
66
00:03:17,960 --> 00:03:23,120
This was built by Felicity, a developer that I met personally at WWDC and it is a replacement
67
00:03:23,120 --> 00:03:28,079
to be able to browse your Apple Music but if you also store music on a Plex server,
68
00:03:28,079 --> 00:03:29,920
You can put it all in one place.
69
00:03:30,000 --> 00:03:34,240
It feels a lot like the stock Apple Music app, really nicely designed.
70
00:03:34,400 --> 00:03:36,840
It's available on iPhone, iPad, Mac and Apple Watch.
71
00:03:37,099 --> 00:03:38,479
Again, free to download.
72
00:03:38,479 --> 00:03:41,240
Try it out, especially if you have a Plex library of music
73
00:03:41,240 --> 00:03:43,719
and you want to be able to browse that and Apple Music together.
74
00:03:43,879 --> 00:03:45,020
And another free app.
75
00:03:45,020 --> 00:03:48,780
I recently moved away from Raycast, Raycast updated to version two.
76
00:03:48,780 --> 00:03:51,560
I was loving Raycast for the last month when I switched to it.
77
00:03:51,879 --> 00:03:55,659
But version two has broken some things like my shortcuts and a lot of controls.
78
00:03:55,900 --> 00:03:58,659
And so I've been experimenting going back to Spotlight.
79
00:03:58,659 --> 00:04:00,219
I created a bunch of shortcuts.
80
00:04:00,000 --> 00:04:04,259
And I'll link my YouTube video explaining how I'm going back to Spotlight and building
81
00:04:04,259 --> 00:04:08,280
all those Raycast features with shortcuts, even things like window management.
82
00:04:08,280 --> 00:04:13,400
But one of the applications that make it possible is something called HyperKey, free macOS app,
83
00:04:13,400 --> 00:04:14,400
free to download.
84
00:04:14,400 --> 00:04:19,319
And what this does, it lets you remap the Caps Lock key on your keyboard to be a bunch
85
00:04:19,319 --> 00:04:20,420
of modifiers.
86
00:04:20,420 --> 00:04:26,120
So I have it so when I hold the Caps Lock key, it actually reads Shift-Option-Control-Command.
87
00:04:26,120 --> 00:04:27,740
Four modifier keys.
88
00:04:27,740 --> 00:04:30,040
This way you can create keyboard shortcuts that
89
00:04:30,000 --> 00:04:33,319
rather than having to hold down all those keys and then a letter you can
90
00:04:33,319 --> 00:04:36,920
just hold the caps lock key and then W. I mapped that to a shortcut that will
91
00:04:36,920 --> 00:04:40,600
take whatever the active app window is put it on the left side of my screen a
92
00:04:40,600 --> 00:04:44,600
quarter size and place it precisely where I want. In my video I talk about a
93
00:04:44,600 --> 00:04:48,240
lot of other use cases I have for the hyper key but it's a great free app
94
00:04:48,240 --> 00:04:52,759
that's linked down below. And finally on something deep MKBHD Marcus Brownlee one
95
00:04:52,759 --> 00:04:56,480
of the largest tech
youtubers on the platform he posted a 12-minute video
96
00:04:56,480 --> 00:05:00,019
commenting on the new YouTube feature that we've mentioned on primary.
97
00:05:00,000 --> 00:05:06,639
the AB video test feature that will be coming next year in 2027. There's lots of AB thumbnail
98
00:05:06,639 --> 00:05:10,319
tests on YouTube right now, you might even notice as you go on YouTube, you see a video
99
00:05:10,319 --> 00:05:13,839
from a creator, you know, the thumbnail and title might look one way. And the next time
100
00:05:13,839 --> 00:05:18,240
you log on to YouTube, or maybe when you look on your phone versus the TV, you see a different
101
00:05:18,240 --> 00:05:21,920
thumbnail. And I do this a lot too, I will test three different thumbnails and three different
102
00:05:21,920 --> 00:05:26,959
titles. And whichever gets the most watch time from people who click through, YouTube will say
103
00:05:26,959 --> 00:05:30,000
that one wins and then make that the thumbnail and title. And I think it's
104
00:05:30,000 --> 00:05:33,139
a great tool to be able to test which is most effective. YouTube is launching
105
00:05:33,139 --> 00:05:38,759
video A/B testing, meaning when you click the same YouTube link versus your friend
106
00:05:38,759 --> 00:05:42,860
or someone else clicking that same link, you might see a different video. Now
107
00:05:42,860 --> 00:05:46,199
Marquez explains that he talked to YouTube and they are going to have an
108
00:05:46,199 --> 00:05:51,680
algorithm that tells how different these videos are. So supposedly you won't be
109
00:05:51,680 --> 00:05:56,480
able to create a video how the new iPhone 18 is amazing and then create
110
00:05:56,480 --> 00:06:02,079
another video the iPhone 18 is terrible. Supposedly this algorithm...
111
00:06:00,000 --> 00:06:04,800
will catch how different those videos are and not allow a creator to A/B test
112
00:06:04,800 --> 00:06:08,879
that. You also can't go backwards and A/B test a video you already published. This
113
00:06:08,879 --> 00:06:12,680
will only be for new videos going on a platform next year. But Marques has a
114
00:06:12,680 --> 00:06:16,399
good point, meaning comments will be broken. So if you comment on a YouTube
115
00:06:16,399 --> 00:06:20,360
video and the timestamp is in the comment, which version of the video did
116
00:06:20,360 --> 00:06:24,079
that person see? Will those timestamps be broken? And you might watch the first 30
117
00:06:24,079 --> 00:06:28,439
seconds of a video, share it to someone else, and they might see a different first
118
00:06:28,439 --> 00:06:29,959
30 seconds of a video.
119
00:06:30,000 --> 00:06:32,100
You're not even talking about the same video.
120
00:06:32,399 --> 00:06:35,399
Again, with this algorithm testing, maybe it will be clear that these
121
00:06:35,399 --> 00:06:40,439
videos are not that different, but it creates this world where the YouTube
122
00:06:40,480 --> 00:06:45,360
video, even at the same link that you may watch might be a different experience
123
00:06:45,360 --> 00:06:50,160
than the video that I watch at the same link and the algorithms that already
124
00:06:50,160 --> 00:06:54,079
power the feeds on Instagram and Facebook and threads and all these social media
125
00:06:54,079 --> 00:06:57,399
platforms, it creates different experiences of the world.
126
00:06:57,600 --> 00:07:00,160
And that's why many of these platforms are echo chain.
127
00:07:01,000 --> 00:07:06,199
the biases that people may have just get confirmed and reaffirmed because the algorithm gives
128
00:07:06,199 --> 00:07:09,160
them the content that they interact with and want to watch.
129
00:07:09,160 --> 00:07:12,240
And while YouTube is based on an algorithm as well, at least the videos you see on the
130
00:07:12,240 --> 00:07:17,920
homepage, this idea that we click the same YouTube link and we might see different videos
131
00:07:17,920 --> 00:07:21,959
at that same destination, that doesn't feel super great.
132
00:07:21,959 --> 00:07:26,519
Now it's also a huge lift for someone like me who's a solo creator to edit two or even
133
00:07:26,519 --> 00:07:29,959
three videos differently and upload that as one to the platform.
134
00:07:30,000 --> 00:07:44,000
So I will probably not be taking advantage of this, but I'm curious if YouTube walks this back or just how this feature can be used, especially in relation to Marquez posting this video, which again, he's very influential on the platform and one of the biggest tech creators out there.
135
00:07:44,000 --> 00:07:48,000
But let me know what you think. How do you feel about A/B video testing on YouTube?
136
00:07:48,000 --> 00:07:53,000
You can sign up for my free newsletter by scanning this QR code. And when you reply to that, I get every email and read those.
137
00:07:53,000 --> 00:07:56,000
So I'd love to hear from you. A/B video testing. How do you feel?
138
00:07:56,000 --> 00:08:00,160
feel. And if you haven't yet, would greatly appreciate a 5-star rating and review on Apple
139
00:08:00,000 --> 00:08:07,000
Podcasts.