Fishkins

You can be concerned about all of that at the same time. I'm a lifelong vegetarian who had never owned a car, but I'm also wary of attempts to alter existing ecosystems. I don't think it should be out of the question, but it should be approached with extreme caution.

> I missed the part where this has anything to do with saving time

It lets you put the non-car stuff closer together, so you're traveling less distance to get to the same place. It requires urban design, not just a single person switching between modes of transit.

(Although switching to cycling can often make transit both faster for you and the people around you in a city because you aren't as affected by traffic and don't create as much traffic)

The boat itself rocks, but do you see the background changing to indicate the boat is progressing through the environment? I only see that in the 3.1 Pro example. I believe that's what the OP meant.

I'm sure most folks run Claude without isolation or sandboxing. It's a terrible idea, but even most professional software developers don't think much about security.

There many decent options (cloud VMs, local VMs, Docker, the built-in sandboxing). My point is just that folks should research and set up at least one of them before running an agent.

Right, that's what I was referring to

Thanks for the data-based comment!

Have you noticed any change in that trend in the past year or two, or is it continuing to get better?

I agree with most of this, with one important exception: you should have some form of sandboxing in place before running any local AI agent. The easiest way to do that is with .claude/settings.json[0].

This is important no matter how experienced you are, but arguable the most important when you don't know what you're doing.

0: or if you don't want to learn about that, you can use Claude Code Web

I do something similar. I leave up and down arrows alone, but have ctrl+p and ctrl+n behave as you describe.

> humans also make mistakes

This is broadly true, but not comparable when you get into any detail. The mistakes current frontier models make are more frequent, more confident, less predictable, and much less consistent than mistakes from any human I'd work with.

IME, all of the QA measures you mention are more difficult and less reliable than understanding things properly and writing correct code from the beginning. For critical production systems, mediocre code has significant negative value to me compared to a fresh start.

There are plenty of net-positive uses for AI. Throwaway prototyping, certain boilerplate migration tasks, or anything that you can easily add automated deterministic checks for that fully covers all of the behavior you care about. Most production systems are complicated enough that those QA techniques are insufficient to determine the code has the properties you need.

I had the same experience as peer comments. I'm on Pixel 8 and Google Fi. When I check for updates, I'm told I'm up-to-date with the last update being over a month old.