Ex-OpenAI Peter Deng says AI may be rewiring how kids think, and education could shift with it.
The skill won't be memorizing answers. It'll be learning how to ask better questions to unlock deeper thinking.
“When the calculator was invented, people didn't stop doing math. They
Dwarkesh says the Age of Scaling isn’t over.
Companies will keep pouring exponentially more compute into training because the value of intelligence is so high, it’s worth even $1 trillion if it gets us to AGI.
Right now, compute usage for frontier models grows ~4x per year,
i’m still not convinced.
how can we achieve AGI by simply scaling up LLMs?
no matter how much training data you give llms they’re still just a pattern prediction system
llms will be a key part of achieving AGI -- but calling llm alone AGI is not enough
we need at least 1 big
great news for everyone building (lots of customers)
great news for everyone using AI (useful products)
great news for everyone making AI models (large cost is worth it)
OpenAI's and Jony Ive's hardware is 'not an in-ear device, nor a wearable device'. I'm still predicting a tiny robot that sits on your desk and hangs upon your every word. Syncs with your Chat account when in wifi range, and runs a small bespoke OAI model locally when not.
ChatGPT's triumphant advance is unstoppable. Still the most important download in the Apple Appstore.
I don't know anyone (apart from grandparents) who doesn't use ChatGPT regularly. AI has reached the masses and it's hard to imagine life without it.
Daniel Kokotajlo says even if AGI weights were open to everyone, true equality won’t last
Those with more GPUs would quickly pull ahead and race toward AGI+, AGI++, and beyond
it’s not exactly winner-takes-all, but scale decides who dominates,
"and that’s part of what makes
our engineering and compute teams do incredible work to rapidly scale to meet customer demand for chatgpt.
a lot of blood sweat and tears go into this, and they make it look relatively easy.
i have never seen a team handle a 2.5 year sprint with such grace!
Google DeepMind just dropped Gemini Robotics on-device model
> runs locally on robots without requiring an internet connection
> model performs at almost at par to cloud-based Gemini model
> outperforms other on-device models
> devs can control and fine-tune the model using
OpenAI is developing the "everything app"
OpenAI plans to create a direct competitor to Google Workspace and Microsoft Office with ChatGPT. Drafts for collaborative functions such as joint document editing and multi-user chats have already existed for over a year.
The first