GPT5 WON'T FALL FOR YOUR CLEVER PROMPTS
"We've also completely overhauled how we do safety training.
In our older models, the system would look at the user prompt and then decide either to outright refuse or fully comply.
That approach works well in most settings, but

GPT-5 is here - and it's #1 across the board.
🥇#1 in Text, WebDev, and Vision Arena
🥇#1 in Hard Prompts, Coding, Math, Creative Writing, and more

gpt-5 is now free in cursor_ai, go try it out! (reload cursor if you don't see it yet)
we've worked closely with a certain company to make this happen

GPT-5 is here - and it's #1 across the board.
#1 in Text, WebDev, and Vision Arena
#1 in Hard Prompts, Coding, Math, Creative Writing

Graphiti simplifies building knowledge graphs with real-time, temporal data support. Perfect for LLM agents and advanced AI applications.

For those not watching the stream - Sam talked at length about how they trained GPT-5 to be better at answering health Qs + talking through issues.
They measure it on a HealthBench that a certain AI company created with 250+ physicians evaluating outputs.

OPENAI TEAM: GPT-5 TURNS INTO A LANGUAGE TUTOR
"I think our model is trying to say: thank you for your patience waiting for GPT-5.
Now, let's move on to something more practical.
Learning is a top use case for our voice model. People love to use it to dive deep into topics,

We're happy to announce that the Nethermind Security team has completed another review for a decentralized finance project. This time, the...

Big news 🚀
The Stylus course launches today!
Stylus is the FUTURE of EVM development. It saves 90%+ on gas fo…

SonicDB removes read amplification by storing state in binary files instead of a key-value database.
On the first 11 mi…

I spent the last 24 hours finetuning Dots.OCR with different datasets. Here are some of the things I learned...

BOOOOM! You can now run certain AI models natively in T4 for FREE! 🔥
Powered by Transformers ⚡
The setup tak…

NEW - Nvidia's chief security officer has published a blog post insisting that its GPUs "do not and should not have kill switches and backdoors." It comes amid pressure from both sides of the Pacific, with some US lawmakers pushing Nvidia to grant the government backdoors to AI

The most pressing issues to be addressed are engineering problems and training stability.

Exactly. If we're gonna trust AI with real decisions, the data behind it has to be solid and trackable. This makes that possible.

Would it be possible to get Grok to be able to tell what's real from false by analyzing patterns and performing first principles logic?

I understand that it should be related to the capacity of the utxo you created.