When national interests require that, it can get a firmware update which sends a copy of data to comrades in U.S. Ministerium für Staatssicherheit even before that e2e encrypted copy reaches your phone.

tried this:

   Generate Earth-like planet (10,000km diameter, 24h rotation, 23° axial tilt).
   Single supercontinent spanning 40% of surface area, centered on equator.
   Remaining 60% is ocean.

   Continent features:
     - Central desert plateau at 1,500m elevation, approximately 2,000km × 1,500km
     - Mountain range along entire eastern continental margin, peaks 4,000-6,000m
     - Gradual elevation decrease toward western coast

Looks like it understood only „earth-like“ and 23° axial tilt, the map is fractal.

Thanks for trying it out. I think I need to change the model that generates the preset. It’s currently using gpt-5-mini.

You can call it “output granularity” and allow Java logger style configuration, e.g. allowing certain operations to be very verbose while others being simply aggregated

If we're going there, we need to make the logging dynamically configurable with Log4J-style JNDI and LDAP. It's entirely secure as history has shown - and no matter what, it'll still be more secure than installing OpenClaw!

(Kidding aside, logging complexity is a slippery slope, and I think it's important, perhaps even at a societal level, for an organization like Anthropic to default to a posture that allows people to feel they have visibility into where their agentic workflows are getting their context from. To the extent that "___ puts you in control" becomes important as rogue agentic behavior is increasingly publicized, it's in keeping with, and arguably critical to, Claude's brand messaging.)

They don’t have to reproduce it literally. It’s an UX problem with many solutions. My point is, you cannot settle on some „average“ solution here. It’s likely that some agents, some operations will be more trustworthy, some less, but that will be highly dependent on context of the execution.

I fully agree with you on almost everything you wrote in this thread, but I’m not sure this is the right answer. I myself currently spend a lot of time with CC and belong to that group of developers who don’t care about this problem. It’s likely that I’m not alone. So it doesn’t have to be the least professional audience they serve with this update. It’s possible that Anthropic knows what are they doing (e.g. reducing level of detail to simplify task of finding something more important in the output) and it’s also possible that they are simply making stupid product decisions because they have a cowboy PM who attacks some OKR screaming yahoo. We don’t know. In the end having multiple verbosity levels configured with granularity similar to java loggers would be nice.

Oh totally - I'm definitely not saying that they made the decision to cater to non-dev users, just that it's a possibility. Totally agree with you that at the end of the day, we haven't the foggiest idea.

> Why would you expect people not living in the Netherlands to just know this?

It’s generally a good idea to learn something about your travel destination and not to assume everything there is the same as where you live. The world is big and diverse.

If AWS becomes a separate business now, they may be able to build better products, given a bit less focus on one large customer (Idk if Amazon.com has any priority in their product roadmaps now)

* Long term cannabis users that do not slide into something worse over time or just do not consume too much.

Yeah, maybe people able to stick to moderation over many years are just more likely to have their lives together in general.

I doubt it’s killing open source. The “too big to fail” software will be maintained no matter what, but the contribution model will change. It is not great, but we can live with it - majority of users of OSS never touch the code, so nothing is going to change for them. For a few enthusiasts the barrier will be higher, but we need some trust building incorporated in the process anyway.

The small libraries will be eliminated as a viable solution for production use, but that’s a good thing. They are supply chain risk, which is significantly amplified in the LLM age.

It may happen and it will be great if it happens, when open training datasets will replace those libraries to recalibrate LLM output and shift it from legacy to more modern approaches, as well as teaching how to achieve certain things.

Privacy by design is easy. If you are incapable of dealing with GDPR, don’t start a company, because you lack survival skills amyway.

The most interesting part is that they do not rely on Western software solutions (Russia still needs hardware, China may reach full autonomy soon enough). If they could do it relatively quickly, EU can do it too. And EU now has exactly the same incentives.