I've written before how it seems every other day there’s an AI startup raising a massive round with unicorn valuations.
Yet, most of these startups don’t create AI — they just implement AI. That’s useful for users, but is neither defensible nor sustainable for the startup.
It does feel like a .com bubble. Everyone is offering AI even if it’s insignificant or silly to have for a particular business. Yet it sounds great in your pitch.
I wonder if high interest rates will drive investors away from utopian investments.
Most AI should be stealth AI. Any business that doesn't incorporate AI in some way is going to get overtaken by one which does — but that doesn't mean it's central to their value proposition, and certainly not central to their freaking pitch!
In your newsletter "3-2-1 Traction: Is AI a bubble? Also, original pitches suck!" that came out some time ago, you write:
"
...Investors are trying to get information as quickly as possible, so that they can digest, evaluate, and compare....
...Yet, most of these startups don’t create AI — they just implement AI. That’s useful for users, but is neither defensible nor sustainable for the startup.
And it really doesn’t make for a good investment.
Sure smells like a bubble to me. ..."
It is not clear to me how investors make a decision without analyzing the current possibilities of technologies.
As an example, the startup www.synthesia.io raised 1 billion. At the same time, it is obvious that open AI technologies allow you to do the same thing, for free.
It seems that investors would rather invest a billion quickly than take the time to understand modern technology.
Thanks for reading! Synthesia is valued at more than $1b, but they only raised about $90m. Against $16m in revenue, that valuation is still... high. 😅
Technology isn't the same as product, though. The vast majority of unicorns aren't using original tech — Uber, AirBnb, etc. Even ByteDance and Stripe didn't create much in the way of new tech. Their value is in their differentiation (originally) and in their user base (eventually).
But your question is a great one! Investors should not make investments without understanding the tech. Typically, that's done in due diligence (either through their own firm, or through people they know or hire to evaluate the deal). But I think the mistake in the AI bubble is a much more critical one — they're ignoring what they hammer on in every other pitch they hear (differentiation, defensibility, etc) simply because the proverbial penny is shiny.
Good point. But in the sea of these AI startups that are just implementing AI and are not really great, there's a couple that really have potential. Sadly, I believe they'll end up in the shadow of these mainstream startups that aren't as good.
There are exceptions to every rule, sure. I was drawing a distinction between startups that leverage AI (which should be every startup, imo) and startups that leverage AI as their sole value prop. For the latter, they better be defensible in some other way, because their IP is outsourced to third party AI providers.
I won I won I won!
https://youtu.be/OItP8-_mjXw
😁 👏🏻
Hopefully the joke was not lost on most of the audience. I am grateful for the shout!
I've written before how it seems every other day there’s an AI startup raising a massive round with unicorn valuations.
Yet, most of these startups don’t create AI — they just implement AI. That’s useful for users, but is neither defensible nor sustainable for the startup.
And it really doesn’t make for a good investment.
Sure smells like a new AI bubble to me.
It does feel like a .com bubble. Everyone is offering AI even if it’s insignificant or silly to have for a particular business. Yet it sounds great in your pitch.
I wonder if high interest rates will drive investors away from utopian investments.
Most AI should be stealth AI. Any business that doesn't incorporate AI in some way is going to get overtaken by one which does — but that doesn't mean it's central to their value proposition, and certainly not central to their freaking pitch!
In your newsletter "3-2-1 Traction: Is AI a bubble? Also, original pitches suck!" that came out some time ago, you write:
"
...Investors are trying to get information as quickly as possible, so that they can digest, evaluate, and compare....
...Yet, most of these startups don’t create AI — they just implement AI. That’s useful for users, but is neither defensible nor sustainable for the startup.
And it really doesn’t make for a good investment.
Sure smells like a bubble to me. ..."
It is not clear to me how investors make a decision without analyzing the current possibilities of technologies.
As an example, the startup www.synthesia.io raised 1 billion. At the same time, it is obvious that open AI technologies allow you to do the same thing, for free.
It seems that investors would rather invest a billion quickly than take the time to understand modern technology.
Thanks for reading! Synthesia is valued at more than $1b, but they only raised about $90m. Against $16m in revenue, that valuation is still... high. 😅
Technology isn't the same as product, though. The vast majority of unicorns aren't using original tech — Uber, AirBnb, etc. Even ByteDance and Stripe didn't create much in the way of new tech. Their value is in their differentiation (originally) and in their user base (eventually).
But your question is a great one! Investors should not make investments without understanding the tech. Typically, that's done in due diligence (either through their own firm, or through people they know or hire to evaluate the deal). But I think the mistake in the AI bubble is a much more critical one — they're ignoring what they hammer on in every other pitch they hear (differentiation, defensibility, etc) simply because the proverbial penny is shiny.
Not a recipe for success...
Good point. But in the sea of these AI startups that are just implementing AI and are not really great, there's a couple that really have potential. Sadly, I believe they'll end up in the shadow of these mainstream startups that aren't as good.
There are exceptions to every rule, sure. I was drawing a distinction between startups that leverage AI (which should be every startup, imo) and startups that leverage AI as their sole value prop. For the latter, they better be defensible in some other way, because their IP is outsourced to third party AI providers.
It’s either a bubble or we all are losing jobs. Only one way to find out 😄