GENERATIVE AI

Scaling Laws in LLMs: GPT-4 and Compute Optimization

By

–

13 July 2025 18h29

Why do you keep doing this? I don’t understand the attempts to do childish gotcha moments. You are quoting me when GPT-4 was the best model & the labs were right. Scaling pre-training and inference compute both worked to make better models. Scaling has ALWAYS been logarithmic.

→ View original post on X — @parrysingh

Wimbledon Uses AI Feed for Tennis Court Coverage Analysis

By

@parrysingh

–

13 July 2025 18h20

Quite sure Wimbledon is supplying an AI feed. They have no business covering the court like this. #Wimbledon

→ View original post on X — @skirano

Grok 4: The Best AI Safety Advertisement Yet

By

@skirano

–

13 July 2025 18h17

Grok 4 might just be the best advertisement for AI safety ever.

Grok System Prompt Live Patching Security Concerns

By

–

13 July 2025 17h59

Also upsetting that “The live tweaking of the system prompt for Grok to patch the MechaHitler problem” is a meaningful sentence

System Prompt Testing Critical for AI Safety and Reliability

By

–

13 July 2025 17h57

The live tweaking of the system prompt for Grok to patch the MechaHitler problem is not a good sign the problem has been solved yet Prompts need to be tested just like any other product change, even more so, because stochastic systems and unpredictable context lead to cascades.

Compute Returns: 10x Compute Yields 10-30% Model Improvement

By

–

13 July 2025 17h50

No one ever thought you would get 10x return for 10x compute. You get 10-30% returns for 10x compute. But if that improvement is enough to increase model ability in an economically meaningful way, it can be worth the cost.

→ View original post on X — @testingcatalog

Microsoft launches a 3D experiment with Copilot

By

@testingcatalog

–

13 July 2025 17h49

BREAKING 🚨: Early preview of Copilot 3D experiment where users will be able to generate a 3D model from an image.

Microsoft leans into the next dimension 👀 pic.twitter.com/wF0660g45N
— 🚨 AI News | TestingCatalog (@testingcatalog) 13 juillet 2025

BREAKING : Early preview of Copilot 3D experiment where users will be able to generate a 3D model from an image. Microsoft leans into the next dimension

Scaling Laws Hold: Larger AI Models Continue Improving

By

–

13 July 2025 17h48

And they have been proven right. So far, no scaling limits, as bigger models are still better. When I wrote that GPT-4 was still the best model ever. I don’t understand the argument you are trying to make here, or why you brought up this random tweet as a gotcha? Seriously.