Hacker News Reader

Comment by pyman | Hacker News Reader

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 million sounds like a lot, but it's not even close to what you need to build and train a real LLM. You need billions. Most people just don't realise that.

See: https://www.businesswire.com/news/home/20240611122778/en/Bui...

View on HN

This is why I think ai is basically the death of startups as we know them. Only big players can even take a swing. No more underdog garage startups, unless you're just downstream getting dorty bath water from the big boys.

Ai all around is purely about consolidation of power and money. It's bad for workers and ultimately probably bad for the startup world and competition more broadly.

I agree. The infra side is dominated by VCs and big players. And the data is in the hands of regulators, who are looking the other way.

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.

These architectures didn't exist last year. The Chinese are innovating thanks to massive government backing, access to talent, and a clear focus on winning the AI race.

Starcoder was released in 2023, by french/american companies, and there were other coding models before it.

That was right around the time this company had a new $250M funding round, so lack of resources to invest in actual AI is a terrible excuse.

StarCoder is an open-source LLM for code, not text. Builder.ai told investors they were building a virtual assistant called "Natasha", not a code assistant.

I'm just telling you what I read online. Builder.ai wasn't competing with GitHub Copilot, Cody, or CodeWhisperer. Those are code assistants for developers. Builder.ai was building a virtual assistant for customers. It meant to "talk" to clients, gather requirements and automate parts of the build process. Very different space.

And like I said before, creating a dedicated, pre-trained foundation model is expensive. Not to mention a full LLM.

The point is that one doesn't need billions to create useful models — there is no reason they would need new foundation models in the first place — and that 'running out of money' is unlikely to have been their main problem, 250M should be more than enough to create an AI website builder.

I get what you are saying, but you're missing the most important piece of the puzzle: the data.

Everyone talks about models and infrastructure, but without the right data, you've got nothing. And that's where the biggest hidden cost is.

According to the company's own website, they were creating the data themselves. Labelled datasets, transcripts, customer conversations, documents, and more. That's where the real money goes. Millions and millions.

Reply on news.ycombinator.com