Error messages can now come with a patch to fix the underlying problem: https://t.co/sFv9uKDqSj
— Greg Brockman (@gdb) 18 mars 2023
Error messages can now come with a patch to fix the underlying problem:
@gdb
-
Error Messages Now Include Automated Patches for Quick Fixes
By
–
-
AI Tools Democratizing Access and Enhancing User Capabilities
By
–
AI tools lowering the barriers to entry and helping people do more (and have a lot more fun):
-
Predicting AI Model Capabilities: A Safety Breakthrough
By
–
Easy to miss in the GPT-4 blog post — accurately predicting model capability, using 1,000x less compute than the real run. A critical aspect of AI safety will be predicting when various capabilities will arrive — still nascent but taking a real step in that direction.
-
Democratized Red Teaming: Building Robust AI Models Through Adversarial Testing
By
–
Democratized red teaming is one reason we deploy these models. Anticipating that over time the stakes will go up a *lot* over time, and having models that are robust to great adversarial pressure will be critical. Also considering starting a bounty program/network of red-teamers!
-
Haskell Becomes Accessible to Non-Math PhDs
By
–
It is finally possible for those without a math PhD to write Haskell:
-
GPT-4 Misuse Mitigations: Critical Safety Work Behind the Scenes
By
–
Very easy to miss the amount of work (and why it’s so important) our team put into mitigations of GPT-4 misuses:
-
OpenAI Credits Available for All Stripe Atlas Companies
By
–
OpenAI credits for all Stripe @atlas companies:
-
GPT-4 Integration with Stripe Documentation Query Tool
By
–
GPT-4 for asking questions to your @stripe documentation: https://t.co/IVlgrGZmWy pic.twitter.com/npe6rYrSPg
— Greg Brockman (@gdb) 15 mars 2023GPT-4 for asking questions to your @stripe documentation: