Introducing AgronomyArena.com! After feedback from Shane Thomas and Rhishi Pethe this new arena style benchmark allows users to ask questions and vote which (random) model performs best. Also includes a price x performance graph.

Optimus Alpha is excellent. One shots everything. First time I have felt AI takeoff is happening. when do they go live?

TL;DR: We built a transformer-based payments foundation model. It works. For years, Stripe has been using machine learning models trained on discrete features (BIN, zip, payment method, etc.) to improve our products for users. And these feature-by-feature efforts have worked

Nice how really good developer ux = really good aix.

"Why did you stop using me?" you look down at the disposable vape laying on the ground next to your desk, which just spoke. "My internal clock shows you last puffed me 3 days, 22 hours, 36 minutes, 12 seconds ago. Why? Is my flavor displeasing to you? I have a pulse mode-"

Using number of tokens generated to assess popularity of AI is like judging a book by how many words it has. Strange that companies highlight how many trillions of tokens per month they generate.

Have you heard of a digital clone? Here is mine: delphi.ai/johnkempf This thing is trained on all the things I have published over the years. You can ask it questions 24/7/365 It is an interesting time, to have technology this good at impersonating someone, and being able

If we had an AI that was considered conscious, and we slowed it to 1 token per year, would we still call it conscious? Would the same AI at 1k tokens per second consider us conscious? What else is potentially conscious in different time scales? Maybe intelligence is everywhere.

Running Opus 4.5 agents in proxmox host is the most fun I have had in a while. Full infra control and long agent chains - it will spend 20 minutes setting up new containers, configuring them, adding code then running and debugging.

Now every time I do something e.g. taxes I wonder to myself “is this the last time I will ever do this”?

x.com/i/article/2030…

Heirloom self improving skills.md file passed down through generations.

I told my girlfriend I was talking with my Hermes agent on Whatsapp. She got really happy and a bit confused and asked me since when I have a Hermes agent? I told her since this morning, at which point she excitedly asked me if I am planning on buying her something.

Kerbal Space Program has hit an all-time high player count.

open.spotify.com/episode/1viBRy… This is one of the best podcast episodes of all time. Dwarkesh Patel should moderate the presidential debates.