Hacker News: dippogriff

New comment by dippogriff in "The AI backlash is only getting started"

dippogriff — Fri, 26 Jun 2026 14:37:11 +0000

If the labs weren't so aggressive with building datacenters in people's backyards, this could've been a different story. People don't like it when pipelines are built in their backyard either.

New comment by dippogriff in "The AI backlash is only getting started"

dippogriff — Fri, 26 Jun 2026 14:29:02 +0000

They tried that a few times and the mistakes have had consequences.

New comment by dippogriff in "KinetIQ Ascend: Toward 100% Reliable Manipulation and Superhuman Speed"

dippogriff — Fri, 26 Jun 2026 14:24:43 +0000

This is excellent! Very useful takeaways. Being able to properly do continuous training in production is key with robotics data being so hard to come by.

New comment by dippogriff in "Fixing Failures in Browser-Use Models: Why More Data Isn't Enough"

dippogriff — Fri, 26 Jun 2026 14:07:51 +0000

Great work showing on how brittle these GUI benchmarks can be! Love the visuals.

I wonder if SFT is the problem here as opposed to the coordinate discretization; what happens with continuous action space?

New comment by dippogriff in "Autodata: An agentic data scientist to create high quality synthetic data"

dippogriff — Thu, 25 Jun 2026 20:55:53 +0000

This is cool. Creative ways to do external verification is the only path to solving training on LLM slop

New comment by dippogriff in "Every match of the 2026 World Cup as a generative poster"

dippogriff — Thu, 25 Jun 2026 07:23:33 +0000

Neat! minor nit - would be nice if the esc button took you back to the list, instead of having the click the X button

New comment by dippogriff in "Why eval startups fail (2025)"

dippogriff — Wed, 24 Jun 2026 18:18:01 +0000

The current way benchmarks are done and are accepted by the community makes for really uninspired work. Until we're willing to break out of this rigid evaluation format prone to crazy overfitting and gaming, talent will move elsewhere. It is kind of a chicken and egg problem though.

New comment by dippogriff in "For Most of the World, Open-Source AI Is the Only Way Forward"

dippogriff — Wed, 24 Jun 2026 17:52:01 +0000

Edge models will get much better after the current insane capex and organic data for pre-training is dried out. But hard to see how the best open source models will ever come close to the best closed ones.

New comment by dippogriff in "The worthlessness of Vitamin D is mildly exaggerated"

dippogriff — Wed, 24 Jun 2026 05:50:50 +0000

Vice versa, the exaggeration of vitamin D is mildly worthless. Some need supplements, some don't.

New comment by dippogriff in "Qwen-AgentWorld: Language World Models for General Agents"

dippogriff — Wed, 24 Jun 2026 05:47:09 +0000

I'm a fan of this direction. For me the most interesting use case for these world models isn't even training, it's verification. If this thing or some idealized version of it can actually reliably simulate state transitions, could you use it to verify an agent's execution path against hard constraints and replace/eclipse LLMs-as-a-judge?