Update: AI training compute used for top LLMs grows by ~5x per year
That’s a doubling every ~5 months, far faster than the 18 months of Moore’s Law.
Notice that this claims that Gemini Ultra used more compute than GPT-4,
but Gemini Ultra is reportedly far worse than GPT-4 at coding. Something doesn’t add up.
Conclusion:
Massive increases in raw LLM ability are coming, far beyond GPT-4 ability, but not quite yet.
Website
That’s a doubling every ~5 months, far faster than the 18 months of Moore’s Law.
Notice that this claims that Gemini Ultra used more compute than GPT-4,
but Gemini Ultra is reportedly far worse than GPT-4 at coding. Something doesn’t add up.
Conclusion:
Massive increases in raw LLM ability are coming, far beyond GPT-4 ability, but not quite yet.
Website
👀7👍1💯1🤣1
ChatGPT can accurately estimate height from photos
“The girls are using ChatGPT to see if men are lying about their height on dating apps”
“Upload 4 pictures, it uses proportions and surroundings to estimate height.”
“I tested it on 10 friends & family members - all estimates were within 1 inch of their real height”
Article
“The girls are using ChatGPT to see if men are lying about their height on dating apps”
“Upload 4 pictures, it uses proportions and surroundings to estimate height.”
“I tested it on 10 friends & family members - all estimates were within 1 inch of their real height”
Article
🤯22🤡18🤣10👍5😇3
Token Cost of GPT-4 level models over time
Cost of 1M tokens has dropped from $180 to $0.75 in ~18 months = 240x cheaper.
— FWIW, none of the cheap ones are quite to the quality of real GPT-4 on coding, the only real job where AI matters right now, and who cares if they’re cheaper when they’re not yet quite good enough to really do the jobs they could.
Industry wasted last 2 years making models cheaper rather than pushing forward the state of the art.
Cost of 1M tokens has dropped from $180 to $0.75 in ~18 months = 240x cheaper.
— FWIW, none of the cheap ones are quite to the quality of real GPT-4 on coding, the only real job where AI matters right now, and who cares if they’re cheaper when they’re not yet quite good enough to really do the jobs they could.
Industry wasted last 2 years making models cheaper rather than pushing forward the state of the art.
💯22👍5🔥2
Tweet
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
🤣36🤡14😱5👍4🤬1😡1
OpenAI launches new ChatGPT model, o1, with reasoning capabilities of a PhD student
“Similar to how a human may think for a long time before responding to a difficult question, o1 uses a chain of thought when attempting to solve a problem. Through reinforcement learning, o1 learns to hone its chain of thought and refine the strategies it uses. It learns to recognize and correct its mistakes. It learns to break down tricky steps into simpler ones. It learns to try a different approach when the current one isn’t working. This process dramatically improves the model’s ability to reason.“
OpenAI Announcement
“Similar to how a human may think for a long time before responding to a difficult question, o1 uses a chain of thought when attempting to solve a problem. Through reinforcement learning, o1 learns to hone its chain of thought and refine the strategies it uses. It learns to recognize and correct its mistakes. It learns to break down tricky steps into simpler ones. It learns to try a different approach when the current one isn’t working. This process dramatically improves the model’s ability to reason.“
OpenAI Announcement
👀16👍9🎉3😈1
Just 30 messages weekly limit for new o1 model
Unsurprised.
O1 uses an obscene amount of resources, as the announcement confirms.
At the same time, this always was the correct way forward.
There are no rich, energy-poor nations.
Just use vastly more compute.
Bitter lesson
Unsurprised.
O1 uses an obscene amount of resources, as the announcement confirms.
At the same time, this always was the correct way forward.
There are no rich, energy-poor nations.
Just use vastly more compute.
Bitter lesson
👍11🤬9💯6😱2🎉1🤡1😈1