Is this matter about publicly available model or private model? For publicly available model like opus 4.6, bad actors can do whatever they want and Anthropic won't know. If this is only about private custom model, designating public model as supply chain risk doesn't make sense as others can use it.

There might be Gandhian/Nelson Mandela way of handling this. Both fought to change system and didn't teach hatredness towards individual a Racist person. Get arrested peacefully.

Others can work for immediate protest for release of the arrested.

Appeal to common values that they also have, and show how they are violating the religious values they profess.

Technically someone can make some app, that can easily help in getting the citizenship proof for an individual.

I am not from USA.

I had both good and bad experience.

Bad with regex or things involving recursion.

Also bad at integration between modules. I have not tried to solve this yet, by giving documentation of both the modules.

Also model used impacted. To understand java code it was great.

I first ask it to generate the detailed prompt by giving a high level prompt. Then use the detailed prompt to execute the task.

Java code size Upto 20k loc is fine. Other wise context becomes big. So you have to do module by module.

I believe to have a discussion maybe someone has to take an open source code example and then say it doesn't work. Other people can then discuss and decide.

Overall happy with gpt5 and claude code.

Edit:updated bad at integration

> Also bad at integration between modules. I have not tried to solve this yet, by giving documentation of both the modules.

You should first draft the interface and roll out coverage with automated tests, and then prompt your way into filling in the implementation. If you just post a vague prompt on how you want multiple modules workinh together, odds are the output might not met implicit constraints.

Recursive improvement without any physical change maybe limited. If any physical change like more gpu or different network configuration is required to experiment and again change to learn from it that might not be easy. Convincing human to do on AGI behalf may not be that simple. There might be multiple path to try and teams may not agree with each other. Specially if the cost of trial is high.

Forcing rich people to invest in low return area might be better strategy. If they know how to make money productive, they can make that area grow faster so that gdp can grow faster. This will ensure rich people money grows at higher rate only if gdp grows. Asking government to manage money is very in efficient.