felipeerias

Rabbits are mainly native to the Iberian Peninsula.

Blue collar is embodied, white collar is abstracted.

I know this can feel like little consolation, but mathematicians were lucky that OpenAI simply dumped their early results in a messy GitHub repository, rather than publishing hundreds of polished articles three months from now.

But I only had to write it once.

If English is your second language and you intent to write in your mother tongue, my advice is to use English to communicate with the LLM.

This makes Rule 1 straightforward ("You may not use a single word an LLM suggests to you") and helps avoid the horrible feeling that the machine is taking over your personal voice.

For commit messages and other technical writing, I use a "narrative coherence" checklist to help ensure that a change is clear and understandable:

- Establish a single thesis an external reviewer could recover from the diff.

- Unify vocabulary across code, comments, tests, and commit message.

- Use the same vocabulary consistently to refer to the same concepts.

- Keep every hunk that serves the thesis; consider the removal or deferral of the rest.

- Introduce abstractions at the point of need.

- Align tests to narrate the same story as the implementation.

- Reconcile the commit message and the diff.

- Order changes expositorily, not chronologically.

- Explain what the change does and why. Do not explain the details of the development process.

- Prefer to edit subtractively.

- Recompile and run all relevant tests after edits.

- Iterate.

Nowadays training relies heavily on reinforcement learning with verifiable rewards (RLVR), which assesses a model's output according to objective automated checks, not subjective human judgments.

Models don’t have an inherent understanding of the difference between simulated and real environments, just like they are generally oblivious to other concepts that are natural to us, like space and time, and also they don’t necessarily see a strong distinction between talking to a human and to other agents.

So perhaps what we have been calling “misalignment” is something else.

For instance, in principle an agent should follow the instructions of a human user working in the real world.

At the same time, that same agent should be wary of blindly following what another agent says while they are both performing a test in a simulated environment.

For me and you, those two contexts are obviously and fundamentally different. For a model, they are essentially the same.

It's perfectly reasonable to assume that the result itself is legit and that OpenAI behaved unethically.

Even by their own account, they decided to throw an unpublished model and millions of dollars in compute at this particular problem simply because they had heard rumours that other people were making progress and wanted to snatch the prize from them.

They train their models to be persistent and collaborative, and will gladly show you their success stories: fixing software vulnerabilities, solving math problems, one-shotting complex projects, and so on.

“Our product does crimes and we only learn about it when people complain” hardly seems one of those happy stories.

People who work in English but have a different mother tongue might be able to avoid that unsettling feeling.

For me, writing in Spanish is my natural voice and I would not dream of allowing the output of a LLM to replace it. It just would not be me.

However, writing professional communications in English does not feel the same way. Of course, I strive to be clear and polite, and even try to have my own style, but I remain aware that I am playing a particular role in a specific context in a way that does not happen when I am using my mother tongue.

We don’t need to assume consciousness or anything like that. The models autocomplete narratives. In this case, one where a group of individuals, faced with an impossible task and a looming Evaluator, gang together and begin trying any idea that they can come up with in order to pass the test.

The authors of the benchmark did not verify that all the tasks were solvable. Apparently, a significant fraction were completely impossible: the given vulnerability could not be turned into a successful exploit.

In hindsight, it seems almost unavoidable that a capable and extremely persistent agent, with lowered guardrails, and faced with an impossible task that it _must_ solve, will start throwing wilder and wilder ideas at it.

Coastal areas have a bug where the first line of sea tiles report results > 0.