Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
AI

Four-week-old AI Startup Raises Record $113.3 Million in European Push 23

A French start-up founded four weeks ago by a trio of former Meta and Google artificial intelligence researchers has raised $113.3 million in Europe's largest-ever seed round. From a report: Mistral AI's first round of financing values the Paris-based concern at $259 million, including the funds raised, according to people close to the company. The record amount raised highlights the growing frenzy surrounding AI and Europe's desire to create a viable alternative to Silicon Valley companies such as Microsoft-backed OpenAI and Google's DeepMind.

"There is a rising awareness of the fact that this technology is transformative and Europe needs to do something about it, both as a regulator, as a customer and an investor," said Arthur Mensch, Mistral's chief executive. The former DeepMind researcher founded the start-up with Timothee Lacroix and Guillaume Lample, who both recently left Meta after working at Facebook's parent company for the past few years. [...] Mistral has yet to develop its first product, and its first few employees started work only days ago. It plans to launch early next year a new "large language model," similar to the "generative AI" system that powers OpenAI's breakout ChatGPT app.
This discussion has been archived. No new comments can be posted.

Four-week-old AI Startup Raises Record $113.3 Million in European Push

Comments Filter:
  • AI is the new crypto. More evidence that this is a distorted market.

    • by Tablizer ( 95088 )

      Dammit, I need to find a way to cash in on all these fads. I tried to with crypto, but timed it wrong and got kicked in the

      • The secret is to not be involved with the actual tech... only be involved in selling it.
        • by narcc ( 412956 )

          Indeed. You never get involved with the fad directly. That's how you end up with garage full of fidget spinners that you can't even sell for cost. Instead, you sell support for other people jumping on the fad.

          Don't be an author, be a publisher. Don't be a seller, be a manufacturer. Don't buy a coin ... become the mint.

          The old crypto is dead. Fortunes were made and lost. BTC, DOGE, ETH are ghosts waiting to be forgotten. The future of crypto is in the next generation of coins, not those clumsy first

    • Really not. The generative AIs actual provide something, completely unlike crypto. Cryptos only use case has long been more crypto. A far better comparison is the dotcom bubble. There's a bubble. AI is going to be overvalued. But that doesn't mean AI isn't going to be revolutionary, just as the internet turned out to be.
    • Can I interest you in my new startup? It leverages blockchain AI. Also I have bootleg DVDs and genuine Roleks watches too. *opens trench coat *
  • You can keep getting richer.
    • You can keep getting richer.

      When asked whether they would like to be rich, most people will reply "yes".

      When asked what their *plan* is for becoming rich, almost no one has one.

      There are 5 general categories to becoming rich in the US, and those are... ?

      Of the 5, the two that are available to everyone is... what?

      The target amount of money that defines "rich" for your lifestyle is... what?

      See following reply for answers.

      • When asked whether they would like to be rich, most people will reply "yes".

        When asked what their *plan* is for becoming rich, almost no one has one.

        Start by taking the "self authoring" course (find it online). That's the only known resource for success that's backed by strong scientific evidence.

        There are 5 general categories to becoming rich in the US, and those are... ?

        Of the 5, the two that are available to everyone is... what?

        General categories, not specific circumstances. So, "inheriting" or "winning the lottery" are circumstances that go into a single category of "random events that you have no real control over". That's one of the five.

        Case 1, Sales: You can make a fortune in sales if you sell the right things. Telecoms equipment might have a single sale of tens of millions of do

    • by ranton ( 36917 )

      It is obviously nice to be in any social class when compared to other classes with less wealth. It is nice to be middle class if you are poor. It is nice to be upper middle class if you are middle class. It is nice to be wealthy if you are upper middle class. It is always easier to increase your income and wealth depending on your current income and wealth you have.

      It's not clear if any of the three founders of Mistral AI were wealthy before they founded this company. They were certainly upper middle class

  • by Eunomion ( 8640039 ) on Tuesday June 13, 2023 @12:24PM (#63598972)
    Because human sociopaths were just the alpha build.
  • I think the EU certainly has the talent to stumble on something which works well, which is to a certain extent a crapshoot. Both the proposed foundation model rules and the likely impact of copyright and private information laws on make EU hostile to training large models though.

  • I wonder what is the business plan that allowed these three relatively young guys to convince people to 'invest' 130 or so million USD into their idea, because that's all it is - an idea. An idea in search of a business plan?

    • by dysmal ( 3361085 )

      The AI will create their business plan!

    • by narcc ( 412956 )

      An idea in search of a business plan?

      Don't be silly. They're looking for just two things: venture capital and an exit.

      We live a particularly stupid time. I've been involved in more than one business that had just one goal: get purchased by one of the big players. Yeah, it's scummy, but you don't find too many startups these days with dreams of building an empire. Besides, if you're just consulting, you're not morally responsible. It's not your company. It's not your employees. As long as the product you're building isn't harmful, you ca

    • They reached out to investors of pets.com
  • Of course, giving 100 million to a startup sounds crazy, but honestly what they’re paying for sounds pretty straightforward. Pay a few AI veterans who have trained LLMs before to train an LLM, but for the EU. Sounds pretty low-risk.
    • by ranton ( 36917 )

      I guess that depends on the cap table. What you describes sounds more like just hiring these three to run a company, which would mean giving them equity stakes closer to a hired C-level executive than a founder. Right now the investors providing $113 million own 44% of the company. Does this mean each founder owns nearly 20%, or are they only getting a few percentage points vested over a few years? (the rest going to new hires I presume) That could make sense even without a solid business plan if these thre

      • by WDot ( 1286728 )
        What makes LLM s expensive is the cost of training: literally on the order of $100 million to run all those GPUs. The theory is well understood, the data is easy to get (or was, until Reddit realized they could charge for it). I imagine the main “knowledge” they bring to the table is the practice of training and deploying models that use so much hardware all at once, from the institutions they came from.
      • I guess that depends on the cap table. What you describes sounds more like just hiring these three to run a company, which would mean giving them equity stakes closer to a hired C-level executive than a founder. Right now the investors providing $113 million own 44% of the company. Does this mean each founder owns nearly 20%, or are they only getting a few percentage points vested over a few years? (the rest going to new hires I presume) That could make sense even without a solid business plan if these three really are rare experts.

        But if each of these three founders just gained $48 million in equity each without providing something already worth $145 million to these investors, it is among the craziest ideas ever. No amount of knowledge in how to create LLMs is worth $50 million. The ability to either do it or at least learn quickly isn't that rare.

        I don't know the level of talent of these particular folks, or their prior executive/startup experience, but it might not be a bad bet.

        Another way to think of it, you're not investing in the idea or even them, you're spending $145m to build a company in that sector and giving them the job of doing it.

        That's why it's so difficult to get the firs Angel/VC, but so easy to get all the rest. Once you have enough money to build the company and run it for a while it's just a question of whether they'll get cash po

  • But which language?

Enzymes are things invented by biologists that explain things which otherwise require harder thinking. -- Jerome Lettvin

Working...