What happens with all the money they collected from investors? Was it all just squandered away? Pocketed?
Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
81–90 of 283 posts
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#82Do folks think that this was utter negligence by the VCs, or just a pump and dump?
There's a more-or-less useful adage in investing: scared money don't make money. Startups will always carry a risk, and VCs are not betting that the company will be asymptotically good, just good enough to make an exit.
And even more than that, just that some company they invest in will be a winner. It's OK for most to fail, as long as one of them does well. So they invest in lots of long-shots.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#83Earlier quoted context omitted.
Not just VCs. Microsoft was an investor too.
Microsoft invested £250M in Inflection AI, £250M in Builder.ai, and has backed several other companies working on LLMs. They’ve been placing strategic bets across the AI space, but only a few of those companies actually had the talent, infrastructure, and funding needed to build real models. The VP of AI was Craig Saunders, the same person who helped create Amazon Alexa. The problem is, they ran out of money. $500 mi…
...and given the massive success of Amazon Alexa...
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#84Earlier quoted context omitted.
Auto pilot is pretty good at driving within the same lane. Just like a plane pilot needs to do the more complicated stuff like take off and landing, one can expect the same with autopilot (just changing lanes and stopping at lights or stop signs or turning). Full self driving is the mislabeled one. Should be “90% self driving”.
Not when that lane is the HOV lane and traffic is going (speed limit + 15mph). You can always spot the white Tesla on 101 doing the speed limit in the HOV lane during rush hour with a mile of cars behind them and a mile of nothing in front.
The driver can set the auto pilot speed to whatever number they want. If they are going the speed limit, it’s because the driver chose to cruise at that max speed.
Full self driving, however, may be limited to the speed limit. I don’t know, since I don’t use it.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#85Isn’t this what they always tell startups to do? Fake it and get product market fit. I recall the stories of task rabbit where the founder was delivering all the meals.
"Fake it till you make it" was about presenting yourself as an established, stable company to overcome objections about using a startup. Things like having a "Customer Support" phone number and e-mail address that just go to the founders, for example. It's fair game if the founders are actually picking up the phone and doing customer support, and it overcomes one objection people might have about using a startup instead of a big company.
Claiming you can do something specific (use AI to do something) and then using humans to do the labor is something else entirely. If you raise money on that, it's just fraud.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#86Earlier quoted context omitted.
> I really wonder how high-profile investors pour hundreds of millions into a co without doing something simple like ordering an app using a burner account. You'd think that if you're investing $1M+, there's budget for at least getting an intern / assistant to do that.
Microsoft invested £250M in Inflection AI, £250M in Builder.ai, and has backed several other companies working on LLMs. They’ve been placing strategic bets across the AI space, but only a few of those companies actually had the talent, infrastructure, and funding needed to build real models. The VP of AI was Craig Saunders, the same person who helped create Amazon Alexa. The problem is, they ran out of money. $500 mi…
DeepSeek cost just over $5M to train. StarCoder cost around $1M, there is no info for Starcoder2 but unlikely to be more than a few million. The idea of spending billions in training is OpenAI trying to build a moat that might not actually exist.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#87Do folks think that this was utter negligence by the VCs, or just a pump and dump?
There's a more-or-less useful adage in investing: scared money don't make money. Startups will always carry a risk, and VCs are not betting that the company will be asymptotically good, just good enough to make an exit.
> VCs are not betting that the company will be asymptotically good, just good enough to make an exit.
This is a misunderstanding of VC investment. Any competent VC expects most of their investments to go to zero. They're hoping a small percent of their investments will make up for the losses. The goal of a decent VC isn't to avoid bad investments so much as it is to make sure they get one good investment. A good investment in AirBnB/Google/Facebook will make up for dozens of speculative bets that go to zero.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#88Earlier quoted context omitted.
I expect the builder.ai story will break into the mainstream via a book / documentary. There are some insane details and it's the first large-scale AI hype failure - which people are hungry to get the details on - and some big names involved. Another failure of dd - I really wonder how high-profile investors pour hundreds of millions into a co without doing something simple like ordering an app using a burner account…
> I really wonder how high-profile investors pour hundreds of millions into a co without doing something simple like ordering an app using a burner account. You'd think that if you're investing $1M+, there's budget for at least getting an intern / assistant to do that.
>You'd think that if you're investing $1M+, there's budget for at least getting an intern / assistant to do that.
Or having an AI Agent do it...
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#89Earlier quoted context omitted.
Or “auto pilot”
Auto pilot is pretty good at driving within the same lane. Just like a plane pilot needs to do the more complicated stuff like take off and landing, one can expect the same with autopilot (just changing lanes and stopping at lights or stop signs or turning). Full self driving is the mislabeled one. Should be “90% self driving”.
> plane pilot needs to do the more complicated stuff like take off and landing
Auto landing test 3 days ago.
Re: Builder.ai Collapses: $1.5B 'AI' Startup Exposed as 'Indians'?
#90AI = Actual Indians, apparently.
Whats the process for hallucinations? Do 1 in 10 of each of the workers have to be tripping on shrooms all shift?