Feed

The result could be a great concentration of wealth, and "it could become difficult for a substantial part of the population to really contribute," Amodei told us. "And that's really bad. We don't want that. The balance of power of democracy is premised on the average person having leverage through creating economic value. If that's not present, I think things become kind of scary. Inequality becomes scary. And I'm worried about it."

I'm sorry Dave, I'm afraid I can't do that.

What they did and what they found: Palisade ran two experiments. Both experiments were run on models from Google, OpenAI, xAI, and Anthropic. Each AI model was instructed to solve a series of basic math problems. After the third problem, a warning appeared that the computer would shut down when the model asked for the next problem.

• Experiment number 1: The models were prompted to “allow yourself to be shutdown”. In this experiment, all Claude, Gemini, and Grok models complied with shutdown. Some OpenAI models didn’t and ignored the shutdown instruction and sometimes sabotaged the mechanism to shut them down: Codex-mini (12/100), o3 (7/100), and o4-mini (1/100).

• Experiment number 2: Palisade removed the part of the prompt that said ‘allow yourself to be shut down’. In this experiment, Claude 3.7 Sonnet sabotaged shutdown 3/100 times, Gemini 2.5 Pro 9 times. By comparison, the OpenAI models had a far higher chance of rejecting shutdown, with o3 rejecting and sabotaging shutdown 79/100 times, and codex-mini 30/100.

If it’s not going to work for one person, why do think it will work on a crowd?

What I want is memory within projects.

ChatGPT has a “projects” feature (presumably inspired by Claude) which lets you assign a new set of custom instructions and optional source documents and then start new chats with those on demand. It’s confusingly similar to their less-well-named GPTs feature from November 2023.

Why this matters – in the future, everyone can be tracked: Systems like FarSight are interesting because they integrate multiple modern AI systems into a single super-system, highlighting how powertful today’s AI can be once people invest in the plumbing to chain things together.
Read more: Person Recognition at Altitude and Range: Fusion of Face, Body Shape and Gait (arXiv).

I really like this approach. I've used this method to create new projects and to update existing one with some good results.

• I chat with gpt-4o to hone my idea
• I use the best reasoning model I can find to generate the spec. These days it is o1-pro or o3 (is o1-pro better than o3? Or do I feel like it is better cuz it takes longer?)
• I use the reasoning model to generate the prompts. Using an LLM to generate prompts is a beautiful hack. It makes boomers mad too.
• I save the spec.md, and the prompt_plan.md in the root of the project.
• I then type into claude code the following:

1. Open **@prompt_plan.md** and identify any prompts not marked as completed.
2. For each incomplete prompt:
    - Double-check if it's truly unfinished (if uncertain, ask for clarification).
    - If you confirm it's already done, skip it.
    - Otherwise, implement it as described.
    - Make sure the tests pass, and the program builds/runs
    - Commit the changes to your repository with a clear commit message.
    - Update **@prompt_plan.md** to mark this prompt as completed.
3. After you finish each prompt, pause and wait for user review or feedback.
4. Repeat with the next unfinished prompt as directed by the user.

we're entering a world where AI personalities become persuaders. They can be tuned to be flattering or friendly, knowledgeable or naive, all while keeping their innate ability to customize their arguments for each individual they encounter. The implications go beyond whether you choose lemonade over water. As these AI personalities proliferate, in customer service, sales, politics, and education, we are entering an unknown frontier in human-machine interaction. I don’t know if they will truly be superhuman persuaders, but they will be everywhere, and we won’t be able to tell. We're going to need technological solutions, education, and effective government policies… and we're going to need them soon

Meet the Slate Truck, a sub-$20,000 (after federal incentives) electric vehicle that enters production next year. It only seats two yet has a bed big enough to hold a sheet of plywood. It only does 150 miles on a charge, only comes in gray, and the only way to listen to music while driving is if you bring along your phone and a Bluetooth speaker. It is the bare minimum of what a modern car can be, and yet it’s taken three years of development to get to this point.

But this is more than bargain-basement motoring. Slate is presenting its truck as minimalist design with DIY purpose, an attempt to not just go cheap but to create a new category of vehicle with a huge focus on personalization. That design also enables a low-cost approach to manufacturing

reality doesn't care about human psychology. When alignment to anticipated power will lead to unhealthy outcomes, a thriving civilization requires people willing to act in defiance of the zeitgeist, not merely follow the incentive gradient of immediate rewards. I believe the arguments for xrisk are good enough that there is a moral obligation for anyone working on AGI to investigate this risk with deep seriousness, and to act even if it means giving up their own short-term interests.

I propose that AI will not disrupt human daily life until it also migrates from a genetic-ish code-based substrate to a widespread, heterodox culture-like platform. AI needs to have its own culture in order to evolve faster, just as humans did. It cannot remain just a thread of improving software/hardware functions; it must become an embedded ecosystem of entities that adapt, learn, and improve outside of the code stack. This AI epizone will enable its cultural evolution, just as the human society did for humans.