BullshitBench: sorry to say but DeepSeek v4 did really badly, towards the bottom of the table, whether it is high or low reasoning.
@petergostev
-
GPT-5.5 Reinforcement Learning Scaling Across Model Sizes
By
–
GPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size.
— Peter Gostev (@petergostev) 26 avril 2026
My assessment of each:
– Low: weird slop
– Medium: kinda… pic.twitter.com/6YCNqPyzcRGPT-5.5 by Reasoning Effort: I've asked it in Codex to create a physics-based visualisation of RL cycles for different sized models (70b, 1t, 10t), to demonstrate how the amount of RL you can do differs by model size. My assessment of each: – Low: weird slop – Medium: kinda
-
GPT-5.5 Creates 360° Immersive Babylon Gardens World
By
–
Creating an immersive Hanging Gardens of Babylon world with 360° GPT-Image-2 & Codex in 1500 images.
— Peter Gostev (@petergostev) 25 avril 2026
I've tasked GPT-5.5 in Codex to construct a whole world that you can walk through 'google street view' style. It took 1,500 of 2:1 images that can be turned into a 360° immersive… pic.twitter.com/ML9d6EgXTzCreating an immersive Hanging Gardens of Babylon world with 360° GPT-Image-2 & Codex in 1500 images. I've tasked GPT-5.5 in Codex to construct a whole world that you can walk through 'google street view' style. It took 1,500 of 2:1 images that can be turned into a 360° immersive
-
GPT-5.5 and Pro Models Underperform on BullshitBench
By
–
BullshitBench: GPT-5.5 and 5.5-Pro update! They did NOT do well – 5.5 about the same level as GPT-5.4 (around 30-35 rank, 45% pushback). GPT-5.5-Pro did WORSE – only about 35% pushback. I must say the Pro result kind of shocked me. This is actually interesting, what this tells
-
ChatGPT criticized for excessive image generation feature
By
–
I would definitely say it is a little trigger happy in ChatGPT to make everything an image
-
Image Generation Creates Beautiful Websites Timeline
By
–
I must say, the timeline where we ended up with nice websites via image generation is pretty funny. Looks beautiful
-
Queue prompts with mid-thinking correction steering
By
–
Just send a prompt and then it queues, then you can also send a steer prompt to correct it mid thinking, there's a button for that
-
DeepSeek v4 Performance Comparison and Visual Generation Capabilities
By
–
If you want to see what DeepSeek v4 is like, there's no better way than looking at my face alongside the dozens of generations I did side by side with other models
-
Managing CPU and RAM consumption from AI agents
By
–
My main use for the terminal is to work out what the hell is killing my CPU/ram. It is typically some playwright, heavy chrome tabs, could be computer use and some other random stuff. Agents are not very good in tidying up after themselves.
