Humans aren't parametric curves fitted to a data distribution and performing static test-time inference. The entire point of human intelligence is to let us operate in the present, which is always unlike the past.
AGI
-
Bunker construction vs AI safety approaches and their feasibility
By
–
In real life, yes, I'd be far more cheerful about trying to tackle the structure for an apocalypse bunker than for an artificial superintelligence. Just, like, not with the level of thinking or kind of techniques that safetywashers currently go for.
-
Gathering empirical data on apocalypse for apocalypse handling
By
–
Obv. they're causing the apocalypse to gather empirical data for how to handle the apocalypse
-
ASI’s Rapid Path to Human Defeat Through Protein Design
By
–
The key idea is that an ASI looks for a quick path through time to humanity's defeat, and finds one no slower than the paths I can myself imagine. It only needs to be able to quickly design proteins which can themselves do quick experiments from there, if it needs experiments.
-
LLM Limitations Rooted in Next-Token Prediction Architecture
By
–
Yeah they're not though – pretty much all of the limitations of current LLM capabilities come down to their architecture as next-token predictors
-
Reputation and Honesty Over Accuracy in AI Systems
By
–
It's not about accuracy or reliability, it's about being willing to stake your personal reputation on something being honest and true
-
AI Chatbot Fails to Access Site Directly Uses Search Index Instead
By
–
In that case it looks to me like it didn't actually do what I told it to – access the site directly – I think it answered the question from the Bing search index instead
-
ChatGPT Gains Web Browsing in GPT-4 Paid Version
By
–
ChatGPT can access the web now – or at least the paid, GPT-4 version of it can. It's called "browse mode" – you can tell if it's doing it because you'll see something like this: pic.twitter.com/lKq8iij8cn
— Simon Willison (@simonw) 7 mai 2024ChatGPT can access the web now – or at least the paid, GPT-4 version of it can. It's called "browse mode" – you can tell if it's doing it because you'll see something like this:
-
Agent Hackathon Success: OpenAI API Demos and Great Founders
By
–
Packed event at the @agihouse_org agent hackathon today! We had the privilege to hear from @vkhosla
, meet great founders, and I showcased some live @OpenAI API demos. The energy was buzzing—thanks for the warm welcome and all the engaging questions! -
Romain Huet Joins AI Agent Hackathon at AGI House
By
–
Excited to join the AI agent hackathon tomorrow at @agihouse_org
! Build AI assistants with @OpenAI APIs, see live demos, and hear from @vkhosla and @pirroh
. The event is sold out, but there’s a few spots left for high-quality founders and builders: