0:00
Founders pay attention 📍 to what Ben just told us about his journey.
Search this video
0:06
First off, he reviewed 80,000 products.
0:10
Big products are complicated.
0:10
They'll have to start stripping that back to be an interface where you can just get into the product straight away.
0:17
We believe that we are going into a completely different way of doing business in many industries.
0:25
It's just people are reluctant to accept change and it's an uncertain change.
0:31
Marshall McLuhan said that "any new innovation tends to copy the way what it's replacing used to be done."
0:40
Does there need to be this big, bulky product or can there just be a sort of text input interface and then stuff happens on the backend?
0:48
The ability to do something truly new is a challenge And I always made it my job to ask the stupid questions first.
0:57
Creatives now are literally gonna be in the driver's seat.
1:01
Really inspirational to see those people building stuff outta nothing.
1:01
Really, just give it a go.
1:05
What have you got to lose That sense of urgency, that sense of play. I love it.
1:10
Founders pay attention 📍 to what Ben just told us about his journey.
1:11
First off, he reviewed 80,000 products.
1:11
Big products are complicated.
1:11
They'll have to start stripping that back to be an interface where you can just get into the product straight away.
1:11
We believe that we are going into a completely different way of doing business in many industries.
1:12
It's just people are reluctant to accept change and it's an uncertain change.
1:12
Marshall McLuhan said that "any new innovation tends to copy the way what it's replacing used to be done."
1:12
Does there need to be this big, bulky product or can there just be a sort of text input interface and then stuff happens on the backend?
1:13
The ability to do something truly new is a challenge And I always made it my job to ask the stupid questions first.
1:13
Creatives now are literally gonna be in the driver's seat.
1:13
Really inspirational to see those people building stuff outta nothing.
1:13
Really, just give it a go.
1:13
What have you got to lose That sense of urgency, that sense of play. I love it.
1:14
Ben Tossell — Opportunities in AI Jim O'Shaughnessy: Well, hello everyone.
1:14
It's Jim O'Shaughnessy with yet another edition of Infinite Loops.
1:14
I have been really looking forward to my next guest.
1:23
Now, spoiler alert, I have talked to Ben quite a bit about his views on AI, and we're going to get into those on the podcast.
1:29
But my guest is Ben Tossell, the proprietor of Ben's Bites, all about AI, which has followed a similar growth curve to AI itself.
1:45
He went from zero subscriptions to 10,000 in eight weeks. Ben, welcome.
1:53
Ben Tossell: Thank you for having me.
1:55
Jim O'Shaughnessy: So obviously I'm an investor in, and the chairman, executive chairman, of Stability AI, which is a truly open AI company. And so I have been...
2:05
They threw me in the deep end of the pool.
2:10
I had a pretty good layman's knowledge of AI, because it was one of my obsessions for a long time. But now I come to...
2:16
I guess you're my Oracle at Delphi, because in Ben's Bites, which I love, as you know, you scour everything AI, and then you curate it.
2:33
So my first question is, did you see this coming?
2:33
Ben Tossell: The AI wave? Jim O'Shaughnessy: Yeah.
2:39
Ben Tossell: Well, I think I just saw stuff happening where other people were using AI.
2:41
And I think my first real experience of the magic of it was actually Midjourney, using the Discord, and using their imagine prompt, and just seeing the community.
2:55
It was like a million people in that Discord at the time. It was in the summer.
3:00
And I just saw that explode, and I was, like everyone else, probably wide-eyed, and thought, there's loads of stuff here.
3:06
You could do this idea, you could do this idea.
3:10
There's a business here, you could do this. And then that's just...
3:10
Where my brain usually takes me is, if there's ideas like that, I just like following my curiosity, which is what led me to do Ben's Bites.
3:19
But yeah, I saw some of this...
3:24
That's certainly not an early look into AI when using Midjourney at middle users, but that's my first real drop the mic moment of, this has changed everything.
3:30
Everything has changed now.
3:36
And now what do I do with that information?
3:36
How do I sit still on this holiday whilst playing with Midjourney, and not think about all of these other things?
3:41
So that's where I dropped my mic and went, right, okay, I've got to do something here. I want to play with it. I want to try it out. What else is out there? What can I do?
3:50
Jim O'Shaughnessy: Very cool.
3:53
And I know your backstory, and it's super interesting.
3:53
So if you wouldn't mind sharing it with our viewers and listeners, how you came up and why you jumped on this so fast I think is implied in your backstory.
4:04
Ben Tossell: Yeah, so my backstory is, I always wanted to be some hotshot CEO, and that has changed over the years, but that's what I wanted to be always.
4:12
And I couldn't code, I still can't code, and I just refused to believe that the only options available to me were learn to code or find someone who can.
4:24
Because that was all I was told over and over and over.
4:29
And I was like, that's just not...
4:29
That can't be the future of how all of this internet stuff works.
4:32
So then I worked at a company called Product Hunt.
4:38
And Product Hunt is a website that is updated daily with new products, new tech products.
4:44
So I was a community manager there, and the reason I got that job was I was just in the Slack groups that were surrounding Product Hunt.
4:48
So anyone who wanted to be a product maker, anyone wanted to build stuff, I was hanging out in those Slacks.
4:53
I was being the annoying person saying hello to everyone.
4:57
I was the, quote, ideas person, just saying, "Hey, anyone want to help me build something?"
5:02
So I did that a lot, and then found my way to Product Hunt because I was recommended by people at Product Hunt, and they said, "This guy's basically being a community manager already.
5:12
Let's hire him as an actual community manager."
5:18
So then I joined there, and I think it was around 2015 I joined.
5:18
And yeah, I was a community manager.
5:23
I was in charge of seeing what products were submitted, and making sure good ones were on the homepage, making sure that people were having good launches, and good launch experiences, and helping people out in the comments, asking questions.
5:37
And I always made it my job to ask the stupid questions first, because I always felt that it was off-putting if you went to a product that you'd liked, and there was lots of technical speaking.
5:50
I was like, oh, I don't know how to ask what language this is built on, and all of that stuff.
5:55
So I would just ask, "That's a really cool color palette.
5:55
Where'd you design that?" And things like that.
5:58
So I'd really try and get the conversation going at a very, very low bar.
6:02
So I was there for, I think, just over two years when Product Hunt was acquired by AngelList, and I was just a victim of everyone around me.
6:10
I was helping everyone with their launches, everyone was building stuff, and I was seeing all these things. I must have...
6:15
I think I counted, not counted, but I figured out it was around 80,000 products I saw in my time at Product Hunt.
6:20
And I just wanted to be like everyone else.
6:25
I just wanted to build the stuff.
6:25
But again, those responses of, "Learn to code or find someone who can," just felt like they couldn't be the options for me, because I was like, well, I don't know if this idea I've got is any good.
6:37
It might be complete trash, and I'd spend 12 months learning to code to be really poor version of this thing.
6:48
I want to just get something up, launch it, see if people use it.
6:48
If they don't close it and move on.
6:54
So at Product Hunt, luckily I saw companies like Webflow, Zapier, Typeform, all of these kind of companies.
7:01
I thought, well, I can actually build a site on this platform.
7:01
I can tie it together with Zapier into this platform.
7:05
I can make it look and feel like a real product.
7:11
So that's what I started doing.
7:11
I started doing that a lot, and I think I launched 50 plus side projects, we'll call them.
7:16
I don't know if you can call them that if they didn't make any money, but that's what I was doing.
7:20
I was just building stuff more for the fun of the process of, can I come up with an idea, build something, and launch it in a weekend, in a day, in half an hour?
7:32
I actually got to the stage where I think I did one which was 12 startups in 24 hours or something.
7:38
I tried to do something, really push it out.
7:38
So that's what I was doing.
7:38
I was doing a lot of that, and people were noticing that I was doing that, and...
7:42
Not paying, pretty much, projects, but they were always asking me, "How are you doing that?
7:47
How did you manage to build..."
7:51
One of my big ones was an Airbnb clone, essentially.
7:51
I built it without code, and that really took off on Twitter, and everyone's like, "How on earth did you do that?"
8:02
So that's where I shifted my focus to, I can just teach people how to do this stuff.
8:02
So I started then looking at, okay, I can record my screen, I can come up with these wacky ideas, I can build stuff.
8:15
Just record my screen and talk my way through it, and then I can sell access to those videos, essentially.
8:19
And that became Makerpad, which I started in January 2019 as a side project that actually made money that time.
8:24
And that was the difference, actually.
8:30
With all of my projects, I didn't have to push that or force that onto anyone.
8:34
That just came quite naturally, and people were just signing up.
8:34
I didn't have to promote it over and over and over again.
8:38
So that was a real sort of "a-ha" moment for me in my product building career.
8:44
You've got to find something that gains momentum without you having to really force it.
8:50
I think we've seen loads of examples of that with AI, which we can get onto.
8:55
So yeah, Makerpad started January, 2019.
8:55
In the September I raised a small amount of funding, and then in March 2021, we were acquired by Zapier.
9:01
So Zapier is where I currently work, and I'm launching this newsletter.
9:08
It's a little side project that...
9:08
It's just really fun for me.
9:08
I just saw from that summer, all of these projects launched, there's so many things to talk about, so many posts being made.
9:20
Opinion pieces, podcasts, everything.
9:20
So I thought, well, I don't really want to think about building another company.
9:25
I don't really want to think about people having to go and start a new ritual, and have to go to a website, and do that every day, and you have to get content to get them to the website, and all of that stuff.
9:38
And I was like, what's a good easy format that I can just curate a few things, like I had been doing forever really.
9:43
So that's where Ben's Bites came from.
9:43
I just thought, well, I'll try it for 30 days.
9:50
I'll see if there's any interest in it.
9:50
And the first two posts actually were more long form writing.
9:57
So I basically stole the playbook from Milk Road, which is a crypto newsletter that's just been acquired, actually.
10:02
And they do a really good job of making lighthearted stories around what's actually happening in crypto, but mine was going to be in AI.
10:10
So I started writing long emails, and I hated it, and I was terrible at it. It was just not...
10:15
It didn't feel right, and I knew I couldn't keep it up.
10:21
So on the third day I switched to more curated news, simple headlines, links, and then I can add my commentary in and around those pieces.
10:29
And then from there it's really, really taken off.
10:34
And yeah, we're now at 13 and a half thousand subscribers. So yeah, it's been...
10:38
I actually have people reaching out saying, "I adore you newsletter. I love it.
10:41
It's helping me do this.
10:41
It's helping me do that."
10:45
And people are finding hundreds and thousands of customers and users for their products from doing it.
10:49
So that's where we are now.
10:49
And as a bit of a longer winded story than I was anticipating.
10:54
But yeah, that's the whole story now.
10:59
Jim O'Shaughnessy: So what I love about that, Ben, is...
10:59
Founders, pay attention to what Ben just told us about his journey.
11:03
It is filled with excellent examples of how you might want to approach some new thing that you're really interested in, but you haven't figured it out quite yet. So what did Ben do?
11:20
Well, first off, he reviewed 80,000 products.
11:28
I'm going to ask you in a minute about what your win rate on was.
11:28
This one's going to win, this one's going to lose.
11:32
But you watched the process and David Bohm, the quantum physicist, said, " If you really want to learn a subject cold, write a book about it."
11:44
And this is that version of that, it sounds like to me.
11:44
Because what I love is your openness to, you know what?
11:53
I don't have to have a fancy deck with a TAM, a total addressable market.
11:58
I'm just going to keep doing things until the market tells me, "Oh, this one's a winner."
12:05
And then that flexibility of mind is something that I find critical in founders, in general.
12:11
We do a lot of investing in seed stage and series A round companies.
12:18
But then the other thing is, and I'm going to always be pitching my own book here, the open nature of how you discovered... You were in a Slack.
12:23
You saw people conversing, thousands.
12:32
You asked a bunch of questions.
12:32
I love the ask the stupid question, because that's my go to. I don't get it.
12:37
Explain it to me like I'm a golden retriever.
12:37
Because you know if you're thinking that there are 10,000 other people thinking it too, and they just won't ask.
12:53
Next up you're going to be telling me that you're going to be doing a step by step how to actually become an entrepreneur.
12:57
So I applaud everything, and letting the market come to you is really interesting.
13:03
It's a super strong signal.
13:03
We are looking at investing in a seed stage as the lead investor right now, and it has all of those characteristics.
13:10
In other words, the market is clearly in love with this platform.
13:16
Their waiting list is 500, and they're being very diligent and very systematic about the way they're bringing people on, and learning from the people they bring on.
13:27
So boy, that's a great list for founders and for investors to think about. On Product Hunt...
13:33
I'm fascinated by these things. So 80,000, wow.
13:38
Did you keep score of which ones you thought were going to be big, and which ones turned out to be big, versus those that you were wrong about?
13:44
Ben Tossell: I'm not sure, actually, on too many winners.
13:46
I'm not sure I really thought about it like that, but there was always...
13:51
I had a name for myself in the end that I was always going to...
13:51
If you were launching a new weather app, I would always be the first one to say, "Why on earth have you done another weather app?
14:01
Why on earth was it another dating app?"
14:06
So it became a meme of itself then.
14:06
But I know that, of the products that I tried and used, and actually still use to this day, there was one called Card, which is C-A double- Ben Tossell: ... still used to this day.
14:15
There's one called Carrd, which is C-A-R-R-D, .
14:15
co, that's a really simple, one-page website builder that I use for almost anything to just spin up a quick page, and get something zipped through into Zapier, zipped through into Typeform, anything like that.
14:33
That really came out, and I think loads of people loved that straight away.
14:33
And then the founder was a Bootstrap-founder for a long time, and raised a small amount of money, maybe two years ago now.
14:42
But I saw that, and thought it was just so simple, and really up my street with, I'd have to load the code and I can do all that stuff.
14:48
There's a few moments where there was products like Meerkat, which was like the ...
14:53
I think it was just before Periscope, is with all the live video.
14:58
And I was like, "I don't get this. I'm not sure. It's not my format."
15:06
And there's lots of things like that, where I'm sure, actually this would be a good exercise for me, I think, is go back and look at any comments that I had made, that said, "Oh, I think this is amazing.
15:14
I'll see how wrong I was then."
15:14
But yeah, there was just so much.
15:14
I think there's such a broad range of things from big companies like Google.
15:21
We'd help them with launches.
15:25
Or there was lots of YC founders and their companies coming out.
15:25
There really was like, yeah, all those years ago, out of all those products, I can't really remember many, other than the ones I still use today.
15:34
Jim O'Shaughnessy: I love that as a heuristic for looking further.
15:36
So if you find a product that is just so slick that you yourself use it continually.
15:46
I always like to think, "If I like this, there's probably going to be a huge number of other people who are going to like it too."
15:53
I also love your idea that I definitely support, is that the products that catch on, and I'm going to shift over to AI a bit, are the ones that don't require coding ability, that are incredibly easy to use.
16:19
And you're seeing more and more of that.
16:19
We're working to that end diligently at Stability AI for our consumer-facing side.
16:26
The foundational model is as a service, which is where we hope to gain a lot of traction.
16:33
That's a different beast entirely.
16:33
But it does seem to me, that the ease-of-use factor, the usefulness-factor is very high.
16:41
Again, you probably, because you scour every, I mean, I love your newsletter as you know, that's why I wanted to meet you.
16:49
So of all of the stuff since you've been writing the letter, what grabbed you the most?
16:56
And then I want to talk a bit about AI adoption, and your thoughts on that.
17:04
But of all of these longtime subscribers, so I know what's in there, you're a great curator by the way, which ones are like, are you like, "Okay, this is really something."
17:15
Ben Tossell: I think, like many people, it would've been ChatGPT at the time because of some of the things you just mentioned.
17:23
Where you are on the interface where you're using the product, that thing is the thing that you are using.
17:29
I think that that is a huge thing, that I don't think enough people really pay attention to, is unless you can get me in five, 10 seconds, it's going to be very difficult to actually get me to stay.
17:41
Again, that very first experience of using a product matters a lot more than I think people really let on.
17:49
So people spend a lot of time on a lovely-looking landing page and explaining all the benefits.
17:54
No one reads anything, no one scrolls down.
17:54
You just got to have, put me straight into that product.
17:59
So there's things like that, which I think, and that's what I really love and I find most interesting about the AI application layers of things, where lots of things are just like a Google search bar, it almost just looks like, "Type in your input here."
18:16
If that's it, you know anyone probably, who has been on the internet, knows to click in there and type something, and something will happen.
18:22
I think a lot of these AI products, why they're so great and easy to grasp, is because you need to tell, I don't need to go and set up a bunch of things in the backend before I actually get into using the product.
18:39
I don't have to set up my account details, I don't have to put in all these things.
18:44
And even now, seeing products built on top of any of these AI models, sometimes they'll make you require an API key before you go in and you're thinking, "Well, that's a drop-off point for a lot of people already.
18:55
You're going to struggle there, and just let me use a thing, and maybe I test it out, and it doesn't work, or there's some test stuff."
19:05
And I think AI Avatar, that wave, Pieter Levels did a really good job of that because he was building in public.
19:11
And then as soon as you went onto the site, you just saw there's pictures of people in different versions. I get that.
19:14
Okay, I understand if I put a picture of me, that's going to happen.
19:19
It's almost like an input form, "Okay, I'll just fill in the form."
19:24
Everything seems to be that kind of experience.
19:24
So I think that we're just going to get more and more of that.
19:31
I think where lots of these big products are complicated, they'll have to start stripping that back to be an interface where you can just get into the product straight away.
19:39
Jim O'Shaughnessy: Great observation.
19:41
And what's so funny about my thoughts about this, is if you just take a cursory look at what really worked in the earlier era, it's all there.
19:46
I mean I'm holding up an iPhone for our listeners, one of the reasons why this took off, was because it was just so damn easy to use.
20:02
I still remember, I got one of the first iPhones, and habit made me look for the directions.
20:09
And I'm like, "They sent me a flawed thing here, because I don't have the user's manual."
20:19
And then you just, boom, in those days it was a button.
20:27
And of course I'd been trained on the iPod for music.
20:27
But the elegance of simplicity implied in the iPad, in the Kindle that you got from Amazon, it's just like, "Oh, of course."
20:46
And this is one of the things that's been hard for me to rectify, because I deal with a lot of super geniuses.
20:52
And I joke that I did that to ensure I was always the stupidest person in the room, so I could ask the dumb questions.
20:58
And one of the things that I have been finding a challenge, is engineers just love to add, "Oh, but look at how cool it would be if we put this and this and this."
21:12
And I'm in your camp sitting there going, "Yeah, you just added five points of exit from the process."
21:19
And so my own view is that a lot of people are using AI without even knowing it.
21:28
So for example, again with the iPhone, the reason that all those big lenses are gone from the cameras of yesterday, is this is AI.
21:35
And people don't understand that they're using it.
21:41
I'm working up a piece on, "It's not terrifying, guys.
21:41
In fact, it's going to help you become vastly more creative, vastly smarter.
21:50
It's going to relieve you of tedious, boring tasks."
21:50
I was reading about the industrial revolution, and if we just collapse it, you could make the statement that over that period, 70% of jobs went extinct, right?
22:04
Because when the industrial revolution started, everyone was a farmer, at least in America.
22:12
So it coincides with the fears that I hear daily, "It's going to steal my job."
22:18
No, if it does steal your job, you didn't like that job anyway.
22:25
What it's going to do, is it's going to make your next job much more exciting, much more interesting by relieving you of the tedium of cut-paste, whatever.
22:31
So I'm obviously, and I've declared, I'm massively bullish.
22:40
I think that we are literally in the beginning phases of the third tech revolution here, and it's being driven by AI.
22:46
I think that the number of use cases are just, you can keep going and you just keep generating them.
22:53
And you know how people used to always say it's Uber, but for X? Ben Tossell: Yeah.
23:01
Jim O'Shaughnessy: I would modify that and say, X is going to be what AI improves next.
23:01
And then you can fill in the X's, right?
23:10
The medical uses are pretty well known.
23:10
The convenience uses are really well known.
23:17
So it's just a very exciting time for, I think, the technology.
23:17
But I'm a huge science-fiction nerd, love science-fiction, read all the classics, et cetera.
23:24
But the downside is science-fiction set us up for Terminator. Thanks James Cameron.
23:32
Do you hear that from people?
23:32
And if so, what do you tell them?
23:38
Ben Tossell: Well, I think it's just people are reluctant to accept change, and it's an uncertain change.
23:40
So they don't know what it actually means.
23:48
The thought where it'll take your job, I don't really take it that seriously, because I'm like, "It'll just take away the boring stuff, the stuff you don't want to do, stuff you don't want to have to do."
23:58
And I think it would just let you have the more high leverage stuff within that job, if that job is still needed.
24:03
I think it's similar with no-code and all of that, I think is almost abstractions of democratizing software development.
24:16
Where you can put into the hands of anyone the ability to build something with software.
24:23
But before, you used to be able to, you'd have to have ones and zeros, then it was you are learning coding languages, and then frameworks, and then there's some no-code stuff.
24:34
Then AI I think will start, it can write lots of code for you.
24:34
I think, it's just when you look back, you don't think, "Well, I would've much preferred to be coding in ones and zeros instead of doing this."
24:44
And it's easy to look back and see those huge advancements of ...
24:51
Whenever we all have self-driving cars, we'll look back and think, "What?
24:56
We used to drive these big metal machines at each other at 70 miles an hour?
24:56
Is that really what we used to do?"
25:01
And of course, anything on a long enough time scale, that stuff will happen.
25:06
But yeah, I don't know too much really about the threat of the Terminator age.
25:06
I don't know, I just can't see that happening.
25:15
I welcome a robot into my house to do chores and do all sorts.
25:22
Well, yeah, why wouldn't I want that?
25:22
Why would I assume that is then going to turn around and strangle me with my own bed sheet, sort of thing? I don't know that ...
25:28
That's not where my mind goes with any of this stuff.
25:33
I don't know if it's just because I'm optimistic about tech, or I just see, I'm trying to look for the good in it anyway.
25:38
I'm not always looking for the bad of it.
25:43
I know that there are bad actors out there that also are looking at AI, and I think that's why we need to push all the good stuff out and make sure it's all as open as possible, so that we can be prepared and have all our answers for what we need to do, and how we need to solve certain issues and everything else.
26:01
But yeah, I think it's just a reluctance to change, and it's an uncertain, I don't really know what that means for me and my future in this job.
26:10
And that's where people come from, it's that sort of fear.
26:10
But when it comes, I don't think it'll be, "Oh, I wish AI wasn't doing all this stuff for me."
26:15
I think it's just going to be your little AI helper to do all this stuff you don't want to do.
26:21
And I think one thing that I really believe will happen, is all these tools you sign up for to do this thing for you, there's a Chrome extension for that, there's a thing for this, there's an app for that, I think we'll just end up having our own versions of those products on our machines.
26:36
I'll say, "Oh, I want to have a Chrome extension that pulls in these things from this website all the time, and can you write the code and deploy it for me on my machine?"
26:49
And then AI can already write the code.
26:54
I don't think there's a too distant future where that code can then be deployed on your machine, and it will do the things that however you work on your laptop and your things and whatever you are doing, I think it'll just alleviate loads of that stuff when you are always constantly copying from this place to this place.
27:11
Or you're going on these tabs to do certain things over and over again, consuming media, all of these kinds of things.
27:15
I think why wouldn't we want a more personal delivery system, so that everyone's got a PA on a machine at a text input away?
27:27
Jim O'Shaughnessy: Yeah, I share that vision as well.
27:31
I'm a huge believer in the dawn of the completely customizable age, and AI is no different.
27:38
Our company that we sold to, Franklin Templeton, O'Shaughnessy Asset Management, the key there was we figured out and built tech over a decade long period, how to offer a platform, which we call Canvas to investment advisors that allowed for scalable customization.
28:00
As we got feedback from our advisors, I mean, they would come on and say things like, "I've never seen my clients so happy."
28:07
And one said to one of our clients, "Once you've seen this, how could you ever buy a packaged product again?
28:16
Because it customizes to your needs and likes."
28:23
I love your idea that that's the way AI is going to be as well.
28:23
It's going to give, you and I may be very different in what we want to achieve, so our desktops will look very different.
28:36
And I think that that's another one of the things that I just find an incredible and very useful tool for humans to join the exponential age, as it were, without a lot of fuss and bother.
28:48
I love it, because it complies with complex adaptive systems, what AI certainly is.
28:56
Most of the emergence of new innovations comes from the bottom.
28:56
It almost never comes from the top.
29:02
So philosophically, we're seeing the collapse of many of the old models, not just in tech, but everywhere.
29:10
Because the tools that we now have available to us, not just AI, but we're talking via Zoom here, and there's tons of incredibly useful tools that you now have available to you.
29:24
So I think that geography, time, space, they've all collapsed.
29:24
And we now basically have brought 90% of people online, who had a difficult time getting there, and now they're finding all these tools and they're creative.
29:37
But it also brings me to, I want to hear about your hackathon, which O'Shaughnessy Ventures sponsored along with you and several others.
29:43
Were you astounded by some of the stuff that they came up with? Ben Tossell: Yes, I was.
29:50
Some of the stuff just looks, it could be a 50 million funded company, that's the level.
29:58
And that's the stuff that I love to see, is that it doesn't need, the requirement for having investment is not because you have to make this product look and work in this wonderful way, because you can do that with one person sat at home for a week, with just a bit of motivation for a hackathon, for example.
30:18
So there's been a few things there that we've just, voting still live as we are recording now.
30:23
So that'll be closing soon.
30:23
And then we'll be sending out the short list of projects, and there's things like YouTube summaries.
30:29
So some of these AI systems haven't been very good at actually pulling in all of the video information, and then being able to summarize that properly.
30:41
And there's someone in the community who's also hosted meetups and stuff with other Ben's Bites' readers.
30:46
He built that and he's constantly building cool stuff.
30:51
I think that's what AI is enabling, is often, it's like the software magic itself.
30:59
It's just you've got to become the conductor of that orchestra, and figure out how you can pull those pieces together and then actually have something really beautiful and musical come out of it.
31:09
But there's been teams of solo people building stuff, and then there's larger teams building all sorts of stuff.
31:15
I think it was 75 projects submitted.
31:15
So it's been really great to just see all of these people come together and building stuff, sharing with each other, teaming up, sharing ideas, and then just saying, "Oh, how about this?
31:25
You maybe should add this."
31:29
It's just a collaborative, I don't know, some sort of big studio of people.
31:34
Just everyone in there is an artist, all creating these things, and sort of vibing off each other on what you've done and what they've done, and bringing that inspiration together.
31:39
So I'd love to do more of these. I really would.
31:44
I think that's just really inspirational to see those people building stuff out of nothing really.
31:48
They just think, "This is a decent enough motivation."
31:48
And I think that's what it takes sometimes, is you need a bit of kick up the backside to just do something. Just give it a go.
31:59
What have you got to lose?
31:59
And it's just you learn how to do something maybe different than you did before.
32:04
So yeah, there's loads of stuff in there that's cool.
32:08
Jim O'Shaughnessy: And we formed O'Shaughnessy Ventures to take advantage of this very phenomenon.
32:12
We believe that we are going into a completely different way of doing business in many industries.
32:19
And not by the way, just fueled by AI, but fueled by our ability to work collaboratively.
32:28
Also another reason why, to me at least, it's obvious that open is always going to be closed, because open invites everyone, it doesn't exclude.
32:37
It doesn't say, "Yeah, no, you're not up to snuff.
32:46
Where is your certificate from coding camp?"
32:46
It just says, "Come on, let's give it a try."
32:52
And so a lot of people are afraid of that, because it's very different that than what we're used to.
32:58
A lot of people are used to, "I work 9:00 to 5:00 and I do what the boss says."
33:05
And I believe that that era actually ended a while ago, but it took the pandemic to accelerate those trends.
33:12
Like, working from home, for example, or working from anywhere.
33:19
When you're looking, I'm in Manhattan right now, and if you drive up Sixth Avenue into Midtown, it's a ghost town, because all of those buildings were office buildings.
33:25
And now I do think that they'll end up being put to much better and more interesting uses.
33:31
But I could see all sorts of collaborative labs taking up that space at a fraction of the cost that the old fancy corporations did.
33:43
So it's like to me, finally, we're treating people like human beings, and not as chattel to some corporation.
33:53
I'm just incredibly happy about it, because that's how change happens. Almost never ...
34:00
I mean, the classic example would be the old Soviet Union's five-year plan. It was all a joke.
34:07
Everyone knew it was a joke, and they would put up fake numbers if they didn't reach their five-year plan.
34:12
And it was top-down, any kind of top-down system, always the case that they're really unstable.
34:19
Now, complex adaptive systems are unstable too, but there are different solutions.
34:25
Let me ask you, of the projects, and I don't want you to name any names, but did you see a pattern for our viewers and listeners to jot down as they're taking notes?
34:37
A pattern of, "Yeah, that really great entries all had X, Y, and Z, whereas the ones, eh, yeah, not so great, all had these problems?"
34:53
Ben Tossell: Well, there's a balance of originality and usefulness that you can almost, they're the two knobs you can tweak, I think.
35:03
The ones I've seen that I think are really impressive, more often are on the unique side.
35:09
I haven't seen that use case of AI being put to work there before, because I've seeing lots of stuff.
35:15
I see lots of Jasper-like products, or a writing tool for using AI, or a image-generation AI avatar, all of those things that you see a lot of the same thing.
35:22
I think if you're going to try and build something there, there's a very slim chance of what you've got as an opening, and building a better product, a better experience, having the data nodes, all of those things you've got to really try and do to actually carve out your own space there.
35:43
Otherwise, you're just a fast follower.
35:43
I found out in AI that the fast followers are so much faster than they were before AI.
35:49
You'd think of a startup launching, and you could have probably another copy of that in six months.
35:54
But I've literally seen companies that, well, products that have launched the next day.
35:59
There's exact same clone of it just called something else.
36:06
And that is happening a lot.
36:06
And then there's things like YouTube summaries for example.
36:13
I know it's a use case that I've tried to solve for, and some of the products I've used can only summarize the first five minutes, which is not actually useful to actually look at a whole YouTube video.
36:24
So it's either you are solving an actual problem that people are like, "Well, I don't know how to train a model.
36:28
What is training a model? How do I do this?"
36:33
And there's one which is a no-code model builder, which you upload a CSV, it has a column that has the prompts.
36:39
And then you have a column which has the answers that you want to give it as examples.
36:44
And then you can run your own queries against your own data set.
36:44
And that's such a straightforward way to look at it.
36:51
It's a CSV, it's just Excel spreadsheet, a Google sheet.
36:56
You put in some data that you want fed into your machine and then you upload it and that's it.
37:03
Because before you try and look at some of these more technical sites, can you thinking or I'm definitely thinking, "How do I train a model?
37:07
What am I actually putting in and where am I putting that?
37:13
And what does that look like for me to actually put that data somewhere?"
37:17
If I had a company with a thousand customer support emails with the question that they've asked, and somewhere where someone has solved that problem, you could dump that quite easily.
37:25
You could probably use something like Zapier to move all that data into a CSV, format it quite nicely, upload that model, then you can train your own AI support chatbot on top of that model.
37:36
All of these things, again, they're becoming more and more accessible for anyone without the technical ability to do those things previously.
37:48
So I think, yeah, originality, it's difficult to do something that someone else is already doing, especially if they've got their first and they've really gained a market share.
38:00
I think the AI avatars example is just the perfect example of that. I think there's...
38:05
I don't know how many there are now.
38:05
There's so many I've stopped including them 'cause I'm just like, "Well every day I'll just end up talking about another one."
38:14
So yeah, I think we've really got to look at...
38:14
Instead of what you were saying is the Uber for X type thing, there's lots of how do we rethink companies or successful companies?
38:21
How do you rethink some of those things within AI first approach?
38:28
How do you actually look at what that means?
38:32
Does there need to be this big bulky product or can there just be a text input interface and then stuff happens on the backend?
38:38
How can we really rethink some of these companies and tools that we use quite often?
38:44
I'm sure there's a bunch of different things that will come out that we'll think, "Oh yeah, that is so obvious.
38:48
How do I not even think of that?"
38:48
And that will happen time and time again, I'm sure.
38:54
Jim O'Shaughnessy: Yeah, Marshall McLuhan said that, "Any new innovation tends to copy the way what it's replacing used to be done."
39:05
So the example he gives is when we invented movies, the first movies were shot, basically they shot them like a play because that's what they knew.
39:12
They knew, "Oh, I want to watch people acting and I guess I go to plays.
39:18
So we'll just put the camera on and film that play."
39:25
And then of course all of the additional use cases emerge again because a bunch of very clever people are like, "What if we did this?"
39:32
And so the thing that I think is that makes it easier for people to adopt more quickly because they see what they knew and then they're like, "Well I don't know, that's kind of boring."
39:49
But then out come the silent films and we all know the history of Hollywood and how it took over the world and all of that.
40:00
I also agree that the ability to do something truly new is a challenge.
40:00
And one of the things that I always look for when I'm looking at the various decks which are now coming in Avalanche!
40:14
form, does this solve people's pain points?
40:14
Like you mentioned, the avatar Lensa is knocking it out of the park.
40:21
They use Stable Diffusion.
40:21
So I'm very pleased with that.
40:21
And again guys, I'm always going to be talking my book when it comes to both open AI and Stable Diffusion in particular.
40:33
But the thing that I wonder about those kinds of apps, I think they're super cool.
40:39
I've used them because they're easy and I sent them to friends because they're easy and I knew that they would get a kick out of it.
40:47
But how long is Lensa without pivoting?
40:47
How long are our avatar apps?
40:55
No matter how good they are, how long are they going to last?
40:55
Do you have any recommendations for...
41:02
I don't know whether you do this or not, I didn't ask you actually when we had our chat before recording.
41:07
Do you get inquiries quite a bit from startups that are saying, "Hey, help us out here.
41:13
What do you think of this?
41:13
What do you think of that?" Ben Tossell: Every day.
41:16
Yeah, I get lots of that.
41:16
I get lots of people saying, "Oh, we're thinking of doing this," or, "Loved your feedback on that."
41:19
And as much as I want to be able to do that all the time, I think part of my sort of job at doing this newsletter is just to almost give an impartial view and not really push something too much and say, "I don't know how to build a really successful AI company.
41:40
I've gotten into this space because I'm just curious and I'm willing to do some work that most people aren't."
41:44
And I think the same was true for Makerpad is it's not because I was a better entrepreneur or better equipped or anything to do that specific thing.
41:56
And I think what you talked on earlier.
42:01
So I'm quite self-aware of the fact that I'm not a hotshot CEO.
42:01
I didn't actually become that and I realized whilst in that sort of journey that I didn't want to do that, I wanted to do building scrappy stuff, being leaning into that really that kind of founder.
42:13
So I don't want to really give people an opinion on, "Oh, I think you should do this as a business strategy."
42:20
And when I was at Product Hunt, I used to fall into this trap and try and think that I was some sort of startup advisor when I was not.
42:30
And I've got definitely some ideas around the product experience and, "Oh, this would be really good feature," and things like that, which I can always do that.
42:39
People have to figure these things out themselves.
42:39
They have to put it in the hands of users without asking them.
42:43
People will tell you if they hate something.
42:43
They definitely will tell you if they hate something.
42:48
And they will definitely tell you if they love something and if they don't say either you've not done anything wild enough or good enough to elicit any of those responses.
42:58
So I think there needs to be a rethink of that approach.
42:58
I think if you have to go around asking for lots of feedback and if you're willing to just take my feedback and change your product from one thing that I've seen, that's also probably the wrong way to do it.
43:08
So I wouldn't advise anyone to do that.
43:14
But please keep sending me stuff because I need to fill out the newsletter.
43:19
And I love seeing everything but yeah.
43:19
Jim O'Shaughnessy: And I think that curation is going to be a hell of an occupation going forward.
43:21
It's already happening now and I don't see any kind of trend that would disrupt it.
43:27
In other words, everyone laments, "Oh, they don't build things like they used to," and blah, blah blah.
43:35
Without thinking about the fact that, "Well no, but the things we're building right now are so much cooler," and the fact that we are being bombarded by information, ideas, et cetera, to the point where we get way over our Shannon limit and just throw up our hands, it's just like, "Ah, there's just too much."
44:03
And in comes the curator and the curation is a skill because you have to have taste and you have to decide, "Yeah, no, if I put this garbage program," whatever it happens to be, "In Ben's Bites, that's going to reflect on me."
44:20
And so curation I think is going to be huge.
44:28
And like anything it'll probably develop, there'll be channels, there'll be companies, there'll be aggregators.
44:33
And those are all really, really great things in my opinion.
44:40
Sort of shifting gears now to the bigger picture of AI.
44:40
So one of the things that I also believe is that by people working together, connected minds, I'm a huge believer that the networks of the future, which will be global, whereas they used to be, where do you live?
44:54
Where'd you go to school?
45:00
What religion are you, et cetera.
45:00
Those are useful, I'm not saying they're not useful, but the networks of the future are going to be people all over the world and you're going to get a ton of diversity of thought and that's where magic happens in my opinion.
45:17
One thing I always add into that though is I think we should also build all sorts of synthetic minds who think very differently than we think.
45:24
And you can use an animal example.
45:24
So I'm fascinated by crows because they're incredibly smart and learning how crows do what they do, dolphins, whales, elephants, those are all the way I look at it different kinds of cognition.
45:44
And I think that we should not only try to model those, I think we should do more research, but I also think we should create all sorts of synthetic minds.
45:50
And some are going to be really slow and do one thing really well.
45:55
Some are going to be bam, lightning fast.
45:55
And when you need it, there it is.
46:04
But as these minds start interacting with human minds, of course that's where all of these great ideas can come from.
46:12
If you look back at some of the greatest discoveries both in medicine but also in industry, a lot of them were mistakes.
46:18
And so this idea that you can a priori beforehand establish, "Here's what we've got to do," I think is missing the point entirely.
46:34
You need to be able to have a flexibility and pay attention.
46:34
The example that is not AI at all, but it's a great example, is 3M, which of course has the sticky note, God knows how many billions of those they've sold. That was a mistake.
46:50
The scientist was looking for a glue that you could not break away the permanent glue. And it goes on and on.
47:02
Penicillin of course being one of the more famous ones, he forgot to clean up his lab and then he went and he's like, "Oh, what's this?"
47:07
And so keeping an open mind to all of these little gifts of mistakes allows you to not only embrace that mistakes are learning opportunities, we have this sort of in our human OS that we don't want to make a mistake.
47:25
If you don't make a mistake, you'll never achieve anything.
47:29
And so in the broader scope of AI, there are all sorts of analogs going back where people get afraid.
47:38
Whenever a new technology comes we are fear-based creatures essentially.
47:45
And we are especially attentive to novel dangers.
47:45
And because of that, we see something new and we're like, "Woo, is that going to hurt me?
47:54
Is that going to be bad for me?"
47:57
And so it's an kind of axiomatic reaction.
47:57
For example, when a radio came, people truly believed that it killed birds because of these mysterious waves that were killing the poor birds.
48:12
And of course the charlatans and hucksters of the world killed a bunch of birds, put them under a radio tower, took a picture, and there's your story.
48:18
Now we've gotten better at that.
48:24
So deepfakes, other malicious uses of AI is something I take very, very seriously.
48:24
We need, for example, when I say we, I'm speaking specifically about Stability AI, but also the larger we of people who are involved in artificial intelligence need to really be quite serious about things like deepfakes.
48:45
So at Stability, we're going to have a...
48:45
I don't know that we're calling it a hackathon, but we're going to have a contest looking for deepfake detection software.
48:58
Stability's putting $100,000 US in the prize.
48:58
O'Shaughnessy Ventures believes this is so important.
49:05
We're going to add $100,000.
49:05
So it'll be a $200,000 US prize.
49:05
And if it meets our hurdle rates and identifies enough deepfakes...
49:13
By the way, it will have to continually evolve because the deepfakes will get better.
49:18
But one of the other things that we're going to do is we're going to give it away for free.
49:22
And because we feel quite strongly about it, we think that removing roadblocks for people who might be hesitant, don't want to engage I think could be very helpful.
49:34
But it also shows that we're not all Panglossian about this.
49:34
It's not, "Oh, AI has come and it's our savior."
49:40
There are a lot of downsides.
49:40
Have you seen any things where you got sent something where you went, "Whoa, what's this?"
49:53
Ben Tossell: I think probably what everyone else has seen, which is that Tom Cruise deep fake on TikTok.
49:54
That was just so like him.
49:59
I was like, it does take you a little while to realize that it is not real.
49:59
But what was interesting about your last point is think of how many times we do something just because that's how our parents do it, our friends do it, or that's just what we're told.
50:13
If we looked at, and I think it's like a compliance thing, "Oh don't need to figure that out.
50:19
It's already been figured out for me.
50:24
I'll just follow that pattern."
50:24
And I think that's where a lot of people fall into a normal job "I don't want to do it."
50:29
It's probably just because they're being a bit more compliant to that construct of doing that job rather than someone like me who isn't well suited to that.
50:40
I want to go out there and play with something, touch something, do this, make something happen and just follow that curiosity.
50:44
And there's something obviously is in my brain, hasn't quite let the compliance thing take over and say, "Okay, I will just sit here and do that thing."
50:56
But I think it's similar and I think it's very exciting where AI could come in here.
50:56
If you think of any education at all, you just think we've all been taught maths in a specific way.
51:01
If we broke that down or if we looked at how chimps learn maths or whatever those things are there different ways that AI could imagine would be better ways to learn certain topics.
51:18
And then if we were taught those things instead of we are taught it in a specific way in the school system.
51:30
And all of that has its own issues I'm sure.
51:30
But it's one of those things I find really interesting.
51:36
I think there's going to be, yeah, I think there'd just be plenty of solo creators, individuals with their own brands in the age of generated content and images and videos and all of this stuff.
51:49
People are going to look to people for the truth rather than assuming, "Oh this company's... This isn't fake." I believe.
51:55
You going to believe a person over a company.
52:00
And I think that there's going to be lots more of this.
52:00
And I think AI is almost going to help this creator economy, the individual brands or solo VCs, the solo creators of whatever newsletter and businesses that they're creating.
52:12
And I think AI will also help people run those businesses and help with all of the tooling and everything else.
52:17
But just, I do think that lots of the things we experience are because that's just how they have been done and no one has come in to really completely change that version of it and say, "What if we took all of this away and we started again and we really started again and we really looked at something different."
52:40
And I think that's where AI has been really good at being able to say, "Well, we can run through all of these data sets and figure out a pattern where you will be able to," because it'll take you 30 years to do that sort of same thing.
52:51
So there's loads of interesting things there.
52:51
I think that AI will play a really big, big part in.
52:56
Jim O'Shaughnessy: Totally agree.
52:59
We at Stability AI came to an agreement with the nation of Malawi where we are responsible for educating a third of their school kids.
53:10
And we took a first principal's attitude, just like you mentioned there.
53:10
And it's considerably, it's amazingly different than the traditional way of educating.
53:16
And while the jury's still out, we have to see how all the kids do.
53:24
One of the things that we are noticing is that by making it more intuitive for the young learner, they not only like learning, you got to fight to take the tablet away from them.
53:37
They love learning so much.
53:37
And the old paradigm, in my opinion, is so broken.
53:44
And so it was a remnant of the industrial age where you wanted people to do just certain things.
53:51
And so you programmed the right answer in school into their head.
53:57
And I think that what we have coming, what's happening now is the beginning of, I mentioned earlier the exponential age.
54:05
We have a series called The Great Reshuffle about everything that's happening all at once.
54:09
And I personally think that it's going to be a golden age because you mentioned something else, creativity.
54:16
Creatives right now are literally going to be in the driver's seat, in my opinion.
54:22
And by the way, when I talk about a creator, you're a creator because you came up with Ben's Bites.
54:28
I'm a creator because I started a new company.
54:33
Scientists doing research, they're creators, doctors looking for better ways to help their patients.
54:37
They're all creators in addition to of course the artists, the sculptors, et cetera.
54:37
And one of the things that I just am so chuffed about is hopefully as we see these different methods and we see them being embraced, like say uneducation by the young kids, well guess what?
54:54
We're going to be able to come up with new ways to really engage adults who were just like, "Eh, I don't want to do that.
55:05
I don't want to read about that."
55:05
And so I think education, I think the medical applications are going to be monstrous.
55:10
I think we're going to reorganize finance as a discipline. The list is endless.
55:16
But the idea that we are entering an age where people who previously had been excluded by dint of maybe where they lived by dint of them not having the communication.
55:32
And if you just watch what people do, I was rereading a Kevin Kelly book, and he made the point in the book that, "If you really want to see what's going to be the future, go to places that are emerging economies."
55:46
And which if you were going to forecast and there was a farmer, let's say with no electricity and no toilet, even, if they get like a thousand dollars, they're going to spend that on electrifying their house or they're going to spend it on a toilet for God's sakes.
56:08
No, they spend it on smartphones and they spend it on smartphones because of the multiple uses that they can put that device to.
56:15
So I'm a huge believer in loving just what is the market telling us?
56:24
And so the idea that if I have a worry, and I want your view on this, I think that we have moved so rapidly and so fast away from the antiquated laws that are governing us, that something's going to have to give.
56:39
It's like you know why in the United States they picked age 65 to begin Social Security, which is a entitlement program?
56:45
Well, because in 1930, whatever age 65 was when you, according to the actuarial tables died.
56:53
And so essentially what the American government was saying, "Congratulations, you old coot, you beat the actuarial tables."
57:05
And yet we still see these systems with their tentacles into the world of today.
57:05
And so one of the things, I'm an investor in a venture capital firm that is specific to Bangladesh, it's called Anchorless Bangladesh.
57:21
And when I'm looking at these entrepreneurs over there and what they're creating, the systems are better than I have right here in the metro New York area.
57:33
And so, I definitely think that there's got to be some kind of showdown where the government realizes we need to upgrade seriously.
57:39
And if you doubt that, look at the old Eastern block countries, like Estonia for example.
57:48
And look at how much more advanced they are in taxation, in policy, in ownership, in LP.
57:53
In everything because they simply embrace the new. What do you think?
58:00
Ben Tossell: I think it's a really tough job for someone to go and convince governments to do anything that the collective probably think is a good idea.
58:08
But I mean, I feel similarly to you. It has to happen. It should happen.
58:19
But I just don't know that any government really has the balls to try and do something that's quite radical and quite different.
58:27
And I think that's why I find thing like Balaji with his network state and just different completely ideas.
58:32
And I know some of them are quite radical or quite out there.
58:37
I just I don't know what needs to happen for people to change systems that have been in place for so long that it is just like, it's a compliance of, yeah, but this is just how we do it.
58:49
And then, there's got to be some version of, yeah, but it's incorrect or it doesn't work now and everything's changed, so what are we going to do about that?
58:55
But I just, I don't know, I've got no faith in any government I don't think. So I don't know that...
59:05
Jim O'Shaughnessy: I mean that seems to me to also be a consequence of their inaction.
59:09
You've never seen the legacy power structures.
59:09
Be they government, be they journalism, be they universities, et cetera.
59:17
They're plummeting in people's views on them.
59:26
And I think this is, in addition to being a consequence, it's the inactivity and stasis.
59:35
We are moving into, I don't know if you're a Dune fan, but the spice must flow.
59:35
And I augment that often and say the data must flow.
59:41
And it's about static states are no longer workable. They just don't work.
59:49
And so, I'm fascinated by where this goes in terms of watching how society at large and those who seek to govern it, work out a solution that people will accept.
1:00:04
Because you certainly don't want to have a situation where, if you do a poll of Brits or you do a poll of EU members or the people who live in those country.
1:00:15
And everybody fucking hates the government. That's not good. That's not a good look.
1:00:15
And so, I think that we'll work it out, but it's not going to be easy.
1:00:24
And so, people sometimes shy away from that.
1:00:28
It's one of those things though that has to happen.
1:00:28
And so, I think that these kind of tools and just watching what people do.
1:00:35
It's like, I was talking to a friend yesterday and he's a chemical engineer.
1:00:41
And he's quite bright, got promoted several times at the company.
1:00:41
And when the lockdown happened, he was delighted.
1:00:47
He figured out, "I hated going to the office."
1:00:53
And it was a two-hour commute.
1:00:53
It was a waste of time in his opinion.
1:00:53
And I think more enlightened companies stuck with the work from anywhere.
1:01:04
But he was telling me that they did a poll.
1:01:09
And it's a big company, like 12 plus thousand workers.
1:01:09
And they did a poll, "Would you want to go back to the office?"
1:01:15
It was one of the most widely complied with polls.
1:01:15
In other words, people actually took it.
1:01:21
The number, 4% said they wanted to go back to the office.
1:01:29
I mean, you can't just say, "Because that's the way we've always done it," anymore.
1:01:29
What will happen because of the flow, because of the ability of people to use all of these new tools, people are just going to say, "All right. See ya.""
1:01:43
And it's like you're going to watch all of these old systems collapsing and it seems so obvious to me.
1:01:48
And you watch these incredibly inauthentic attempts to say, "But, but, but."
1:01:57
Well, you don't really have much to say.
1:02:04
What would you like to see happen over the next five years?
1:02:04
And back to specifically AI.
1:02:04
What do you think would be, as a curator who I love and I've discovered tons of stuff through your letter, what would you say, "Yeah, if they could do this, that would be super, super cool"?
1:02:26
Ben Tossell: I really, I just want to see new ways of doing things.
1:02:28
Everything seems the same.
1:02:28
But everything is a version of the same product it feels.
1:02:33
Everything's just like we're exactly like them, but just slightly different.
1:02:38
Or we're exactly like them, but our brand color's yellow, or whatever it is.
1:02:44
I'd love to just see that real swings at something completely different.
1:02:44
Like looking at education and thinking, "Okay.
1:02:48
Well, there's a new school that is all developed around AI and everything is self-serve.
1:02:55
And it's all you build on your own and it's all creative and you're touching things.
1:03:01
You're exploring things on your own time and what interests you."
1:03:01
And things like that, I think we will see big swings like that.
1:03:05
And I think there's going to be AI in mental health and there's lots of delicate situations there of course.
1:03:12
But I think the point would be to give everyone access or easier access or feel like the barrier of going to see a therapist is not as astronomically high as it feels like it is for many, many people who actually would really benefit from talking to someone.
1:03:32
Having a chatbot that you can just say, "I'm not feeling very good," and it's talking back to you and it's showing empathy or anything like that is something, there's got to be a bigger thing, a bigger swing that is happening in that space.
1:03:47
And I think there's just every space I want to see disrupted or at least having an AI system really have a go at taking down the old system.
1:03:53
Saying, "You don't need to do it like that anymore. Try it this way."
1:04:00
And I'd love to see how they stood up and see how they actually, what are the differences?
1:04:05
And do we actually use the old system because it is better?
1:04:09
I'm sure in some cases that would be correct, but I'd just love to see those battles I suppose.
1:04:09
That would be most fun for me.
1:04:14
Jim O'Shaughnessy: Yeah, I agree.
1:04:16
We, at Stability AI, do have a mental health chatbot, that when I looked at the number of people who use it, I was dumbstruck.
1:04:23
Clearly, like you with your product, you were like, "Oh, everyone loves this one.
1:04:31
This is the one I'm going to go with."
1:04:34
But taking the feedback, I think you're absolutely right.
1:04:34
And it's part of the reason that I feel like the reason I'm so excited about all this is that the way I view it is it is empowering the individual.
1:04:45
It is empowering society to create new connections, to create new industries.
1:04:52
I don't know what the top jobs of 2045 are, because my guess is most of the tech hasn't been invented yet.
1:05:00
And the way I think about it is, when I see on social media, who's going to be the next Apple, who's going to be the next Google?
1:05:08
My response is, I have no idea, but my guess is they're sitting in a pitch deck on a BC's desk right now.
1:05:22
And that's why this is like a cambrian explosion of all of this stuff.
1:05:22
And we're just really lucky to be alive during this and being part of it.
1:05:31
What do you think's next for you?
1:05:31
What would you like to be working on if we chat three years from now?
1:05:39
Ben Tossell: Oh, I've never been able to plan more than three weeks ahead. So who knows?
1:05:48
Jim O'Shaughnessy: All right, I'll change the question.
1:05:49
Would you want to be working on anything different three weeks from now than you are right now?
1:05:53
Ben Tossell: No, definitely not.
1:05:53
That's why a lot of what I'm doing is an extension of me.
1:05:55
I just want to follow my curiosity and do something that provides someone else value.
1:06:02
If I found it valuable to create all those things myself, someone else might feel the same.
1:06:07
I don't like planning things. I think so much changes.
1:06:11
And AI could come and take my whole newsletter tomorrow, which is fine.
1:06:11
I would embrace that and say, "I'm here for change.
1:06:17
Whatever's next, let's move on to the next thing."
1:06:21
And I just, I like living that way and thinking, "If something comes along, I'll figure it out."
1:06:27
But I really like what we're doing in the newsletter.
1:06:27
Zapier, where I'm working now, has got lots of AI adjacent things.
1:06:32
They've got the open AI integration and a few more integrations coming out that I think again, helps more people build stuff with AI.
1:06:43
And I just think it puts it into the hands of people that can really push it and see where we can take it.
1:06:49
As long as I'm playing around and finding interesting things, whatever I'll be doing, that is where I'll enjoy my time.
1:06:56
Jim O'Shaughnessy: I'll give you a compliment here.
1:07:00
I'm a big fan of Daoism and one of the things about Daoism is this concept of wu wei, actionless activity or being in the flow.
1:07:06
And when I listen to you, what I think is so cool about it is that's what you are. You're in the flow.
1:07:15
You are reacting to the...
1:07:21
It's not like you have...
1:07:21
People often get conflicted. Here's where I want.
1:07:21
I never asked that question when I hired somebody.
1:07:28
Where do you see yourself in five years?
1:07:28
I wanted to try it on you because I thought you might give me the answer you gave me.
1:07:32
And I love that because if you're open-minded, if you are curious, I am finding as I build algorithm and I can't help it.
1:07:38
I'm a quantum and that's what I do in my head.
1:07:45
But everything that I see is that creativity is getting a massive upgrade with these new tools.
1:07:52
And not just creativity, curiosity.
1:07:52
If you are a obsessive curious person, like I am, like you are.
1:08:04
Boy, man, that is going to be such a huge, almost asymmetric opportunity in the new world that we're going in.
1:08:11
And so I'm trying to figure out ways to, how do you get people who are not naturally curious to be curious?
1:08:17
Do you have any thoughts on that?
1:08:23
Ben Tossell: Yeah, I think AI will actually help a lot with this, where you look at all the people using ChatGPT, for example, and all the different use cases that came out of that.
1:08:30
Something that sometimes I still would stare at a blank ChatGPT interface and be like, "I don't know what question I want to ask it next."
1:08:41
But you see all these other people just all sorts of different things.
1:08:47
And I think being able to poke through that and follow a reference down and do all of that, as long as it's in one...
1:08:51
As long as not all the effort is on the person, you can easily follow something.
1:08:57
I think that's where, and that's why a lot of these companies have social products or social interactions just so you can see the most liked model on this platform.
1:09:09
You can go and play with that one.
1:09:09
Oh, actually, let's just toggle some of these back and forth.
1:09:13
But you do need to play with stuff.
1:09:13
And I think that's, we just need people toying with these toys and projects and trying to play around with stuff.
1:09:21
But it doesn't mean that you need to build them.
1:09:25
I don't think that everyone should be looking to build them.
1:09:29
I think just being curious.
1:09:29
And like I said before, I just want to be optimistic because what's the point just sitting here looking at AI and going, "Oh, it's going to take everything away.
1:09:39
What a awful, awful future we've got."
1:09:39
That's not really what's going to happen, is it?
1:09:39
I may as well embrace what a new thing is happening and look at the positives.
1:09:45
And think, oh, that's really interesting.
1:09:50
What if it was doing this?
1:09:50
And asking follow-up questions.
1:09:50
I think that's such a better place and a more interesting place to be in.
1:09:55
Why wouldn't I want to be in that scenario, in that camp?
1:10:00
Rather than looking at it thinking, "Well, I don't want it taking my job and doing my laundry," and all of the rest of it.
1:10:08
I think just, yeah, got to be optimistic about it.
1:10:12
Jim O'Shaughnessy: I think we are brothers from different mothers because that's the way that I look at.
1:10:15
I guess I'm old enough to be your father, so I can't confirm or deny where I was. What year were you born? Ben Tossell: 1990. Jim O'Shaughnessy: Okay.
1:10:25
I will not confirm nor deny where I was in 1989.
1:10:32
Well, this has been awesome. I love your enthusiasm. I love your curation.
1:10:32
I think you're hitting it just right. It's a joy.
1:10:38
Actually, yours is one of the most looked forward to letters.
1:10:44
And I take a lot, as many people know.
1:10:44
I almost always open yours first because it's just filled with such great, wonderful things.
1:10:50
And I, like you, think that the future is extremely bright.
1:10:56
Why don't we do more to tell that story?
1:10:56
And in a way, that's what you're doing with Ben's Bites.
1:11:06
At the end of each of these podcasts, I can never believe how quickly an hour and 25 minutes goes by.
1:11:12
I always ask my guest, we're going to make you and you could take an AI swing on this if you want, but we're going to let you incept the entire population of the globe.
1:11:24
We're going to give you a maybe AI powered device that you can speak into. Here we go.
1:11:30
The AI will translate it into every known language in the world and it will also get it into people's minds.
1:11:38
They're going to wake up the next day and they're going to think that these two new ideas that they just had when they woke up, they're going to think, "That's the best idea I've ever had. I'm going to do that."
1:11:51
What two ideas are you going to incept in the world's population?
1:11:56
Ben Tossell: Gosh, I would probably try and do some sort of you, everyone is now in a competition with each other.
1:12:06
Let's see who can do something more impressive than the other.
1:12:06
And you have to build something with technology to do that.
1:12:14
It's just a global competition.
1:12:14
And yeah, we're just less of the... What are the...
1:12:20
I can't remember the movies where they're like... Oh, the Hunger Games. Less of that. Jim O'Shaughnessy: Yeah. Yes, yes.
1:12:26
Ben Tossell: Less of the Hunger Games, but more of the, what can we do to build interesting things?
1:12:28
And just have some sort of competition around it.
1:12:32
Jim O'Shaughnessy: Which also leads often to cooperation, right?
1:12:35
Ben Tossell: Yeah, because you decide I can't do it. Should we team up? Should we do that?
1:12:40
Or actually, they're doing something similar.
1:12:44
Should we drop this and do something else?
1:12:44
But I think that's how I view the world anyway.
1:12:48
That's how I'm trying to do stuff is, "Oh, I don't really want to do the same thing as anyone else. I'll do something.
1:12:53
I can copy that model but do it slightly differently and do it in my own twist."
1:12:57
Everything has to feel like play to me for me to enjoy it and me to want to do it and be able to do it for an intense period of time, whether that's daily, weekly, for years or whatever it is.
1:13:07
I think that's just what, yeah, I always want it to be like I build for the fun of it rather than for the end goal.
1:13:13
And it always just seems to lead me in good places, into good conversations and everything else.
1:13:20
I will try and somehow get that around a bit more. Jim O'Shaughnessy: Okay. You got one more.
1:13:29
I love the idea about play, by the way.
1:13:29
That's the way I look at the world too. Ben Tossell: What else?
1:13:34
This is just like something that everyone will believe when they wake up. Jim O'Shaughnessy: Yeah.
1:13:40
Ben Tossell: Is that correct? Jim O'Shaughnessy: Yeah.
1:13:42
Ben Tossell: Well, I guess a good one would be then you will be dead in 24 hours.
1:13:50
Do with that what you will.
1:13:50
But again, I'm worried that also becomes the Hunger Games. Jim O'Shaughnessy: Yeah. Yeah.
1:13:54
That's an interesting thought though because that is memento mori, "Remember you will die."
1:13:57
And that can make you do a lot of things more urgently than worrying about what other people are going to think about you or whatever.
1:14:07
That sense of urgency, that sense of play. I love it. Those are great.
1:14:14
Well, listen, Ben, I am going to watch you very closely.
1:14:14
I am absolutely fascinated by what you do. I love your letter.
1:14:20
Where can folks find you? Ben Tossell: Yeah.
1:14:23
I'm on Twitter, @bentossell, T-O-S-S-E-L-L, and then bensbites. co.
1:14:23
That's where the newsletter is and it's a free email newsletter so anyone can sign up. Yeah, take it for spin.
1:14:35
Jim O'Shaughnessy: Easy to use.
1:14:35
Your applying all your own criteria for your own letter.
1:14:39
Well, I'm delighted with the success that you have enjoyed.
1:14:39
One last thought would be, so we're recording this on December 20th.
1:14:44
It's probably going to drop in the first week of January.
1:14:51
And I can't even guess how many new AI innovations are going to be between then and now, which is why the eternal now is so fascinating to me.
1:14:59
And why as long as you want this job, you're going to have this job. Ben Tossell: Yeah, yeah.
1:15:10
Jim O'Shaughnessy: Thank you so much for joining us, Ben, and cheers.
1:15:14
Ben Tossell: Yeah, I appreciate that. Cheers.