Mira Murati Is Launching Her OpenAI Rival: Thinking Machines Lab (theverge.com) 18
Former OpenAI CTO Mira Murati has launched Thinking Machines Lab with several leaders from OpenAI on board, including John Schulman, Barrett Zoph, and Jonathan Lachman. Their mission is "to make AI systems more widely understood, customizable, and generally capable," with a commitment to publishing technical research and code. The Verge reports: In a press release shared with The Verge, the company suggests that it's building products that help humans work with AI, rather than fully autonomous systems. "We're building a future where everyone has access to the knowledge and tools to make AI work for their unique needs and goals," says the press release.
How about (Score:2)
They come up with a use-case first
Re: (Score:2)
Your job. That's their use-case.
Re: (Score:3)
Their use case is to get a piece of the trillion dollar pie. Won't actually be useful or accurate, but it will definitely make them significantly wealthier than they already are.
Resurrecting Thinking Machines Trademark? (Score:5, Interesting)
Thinking Machines Corporation was a supercomputer and AI company from 1983 to 1994. The history in Wikipedia looks like Oracle ended up with everything after it went through Sun and a spinoff also acquired by Oracle. But a quick trademark search makes me think the trademark on Thinking Machines is available for them to use. Is this a good use for a dead brand/trademark?
Re: (Score:2, Interesting)
The trademark was cancelled for failure to renew. That simply means that there is no longer an official registration for it. One would still need to check and see if the name is in use anywhere else in the same industry or category.
TM registration generally give the registrar the ability to ask for enhanced damages if another party if found guilty of infringing. Treble damages if memory serves (IANAL). They also act as an official "time stamp" on the use of the mark/name. But Trademarks also gain some o
Re: (Score:3)
Who would have thought that the real path to
Re: (Score:3)
They made some of the coolest looking machines, the CM-2 and CM-5. Even if the LEDs are eye candy, they still look cool. Plus having processors connected in a 20-dimensional hypercube is a neat trick.
Re: (Score:3)
It's gotta be intentional. Mira should know what role Thinking Machines played in computing and AI in particular.
Why would anybody start a new AI company? (Score:2)
Market is already saturated, they keep getting undercut by some smaller, cheaper team, and every single report says it's just not worth the cost and it has an overall negative impact on everyone that uses it. AI winter is coming, and it's going to be brutal. I mean, they'll still get a cut of the trillion dollars being thrown at AI, because if there's even a chance big business can get rid of employees, err, I mean "improved efficiencies of operations", then they're going to throw as much money of it as the
Re: (Score:2)
You seem a little bitter. But as a fellow human I can understand the sentiment.
I'm pretty convinced we are going to get a substantial amount of AI blended in with our lives over the next few years whether we like it or not. Right now there is a land race for dominance. There are barriers to entry but plenty of room for the newcomers if they can get something rolling.
Re: (Score:2)
While the bubble bursts and you're getting investor money for your bullshit, why shouldn't you?
During the first dot-com you could land a monthly salary that was close to the yearly income of the upper part of middle class, and it was pocket money compared to the stock you got, which usually sent you up higher and higher by the millions at every funding round.
It is the same thing now, only with two or three zeros stacked at the end of the numbers.
Taxes are way too low, the multipliers from "financial innovat
Re: (Score:3)
Sometimes I regret not dropping out of highschool to become a dotcom's token high school dropout whiz kid. But most of the time I'm pretty happy where I am living a quiet life where I am now.
thinkingmachines.ai (Score:1)
Just providing the link:
https://thinkingmachines.ai/ [thinkingmachines.ai]
It was buried in my search results.
Seems nothing of substance other than recognisable names yet.
Interpretability (explainability) (Score:1)
There's a gap in the market for Interpretability and diagnostics, but it's smaller and seems like downstream than just blazing ahead.
They've mentioned interpretability in a word salad with other goals, so there doesn't seem to be much of a strategy yet.
Probably best to think about the personalities involved and judge it from there, or at least "Judge them from their fruits" rather than predict
Allow me to translate (Score:2)
"We're building a future where everyone has access to the knowledge and tools to make AI work for their unique needs and goals," says the press release.
I spent years working directly with a marketing team. Yes, I still wake up in cold sweats, nearly a decade on. But, I can translate this bit of marketing speak. "We building a future where everyone wants to use our AI because we've referred to as a tool rather than a replacement. You can use it while it gathers information, then we'll pull the rug on replacement once it's gathered all the knowledge it can from the userbase."
I'll give 'em this, the newer players in the AI market are getting slightly more int
Good grief! (Score:2)
Naming it after the machines in the dune series that destroyed civilization at a point they had to ban the concept entirely