Sources: OpenAI, Google, and Anthropic are all seeing diminishing returns from costly efforts to build new AI models; a new Gemini model misses internal targets
Three of the leading artificial intelligence companies are seeing diminishing returns from their costly efforts to develop newer models.
Bloomberg
Related Coverage
- Apple's hitting its AI stride right as competition is slowing 9to5Mac · Ryan Christoffel
- AI Companies Reportedly Struggling to Improve Latest Models MacRumors · Tim Hardwick
- OpenAI-President Greg Brockman Returns in a Period of Transition and Turmoil WinBuzzer · Markus Kasanmascheff
- There have been multiple reports that OpenAI's next flagship model codenamed Orion will not perform significantly better than GPT-4. — The issue seems to be that the approach of scaling up AI models by feeding them more data and compute is plateauing especially since they've gobbled up almost all the easily accessible data on the planet. … @carnage4life@mas.to · Dare Obasanjo
- AI Companies Reportedly Struggling to Improve Latest Models MacRumors Forums
Discussion
-
@paulroetzer
Paul Roetzer
on x
Lots of good information here. “After years of pushing out increasingly sophisticated AI products at a breakneck pace, three of the leading AI companies are now seeing diminishing returns from their costly efforts to build newer models.”
-
@basedjensen
@basedjensen
on x
Sigh you know what go ahead run with this wall theory it just gives us so much more room to wipe the floor with ‘omg scaling will not work’ people
-
@btibor91
Tibor Blaho
on x
Bloomberg reports that OpenAI, Google, and Anthropic are having trouble making better models, seeing less progress despite higher costs - Two people familiar with OpenAI's Orion project say the September 2024 model fell short when trying to answer coding questions it hadn't been …
-
@binarybits
Timothy B. Lee
on x
Well done and important story today from Bloomberg about transformer “scaling laws” seeming to run out of steam. https://www.bloomberg.com/... [image]
-
@chehendriksen
Christian Hendriksen
on x
This would be a clear explanation of why OpenAI are prioritizing o1 and the test time compute paradigm.
-
@rachelmetz
Rachel Metz
on x
This is big: we learned 3 of the hottest AI companies — OpenAI, Google and Anthropic — are all struggling to build their next frontier AI models. Read @shiringhaffary, @dinabass, @byJuliaLove and I on what's happening and what's next. Here's a gift link! https://www.bloomberg.com…
-
@morqon
Morgan
on x
another lab struggles: gemini 2.0 “is not living up to internal expectations” [image]
-
@joshm
Josh Miller
on x
Idk if these rumors are true but, if they are, even more reaosn that 2025 will need to be the year that AI *interfaces* help LLMs cross the chasm. Existing frontier models are already so powerful & underutilized. @cursor_ai is the only (truly) AI-native product I've seen. [image]
-
@modestproposal1
@modestproposal1
on x
“After training it, Anthropic found 3.5 Opus performed better on evaluations than the older version but not by as much as it should, given the size of the model and how costly it was to build and run, one of the people said.” [image]