It’s not meaningless - it just shouldn’t be held up as the only thing. Sometimes having a couple proxies is Ok as long as you also look at value in other ways. /shrug

This is the intention. They do not want folks that can’t pay to use their service.

SOTA models cost SOTA prices. Nothing new there

Man I loved programming TI-82s. So many fun ways to build things. I really didn't learn much math that year - I was too enthralled with writing programs to answer the problems for me.

It’s not a need - it’s a fun new thing - fun to see what’s possible and how it helps.

OpenClaw is not easy to set up or user friendly for most (BlueBubbles and Claw had an annoying bug recently) - but the way I have seen it work well requires an up front time investment and then interest compounds RAPIDLY to help manage things and be more productive.

My guess is maybe you’ve never had an assistant or tried a Claw instance? I’ve never had a human assistant but man I’ve had folks that took silly things off my plate and it’s worth it.

LLMs have earned their place in many jobs, but I struggle to see claws as more than a rather expensive waste of time and tokens. Downsides are gargantuan, effects of dead internet theory will be ubiquitous.

Maybe? But I guess I won't find out until I try it a bit.

For now, I'm not posting anything - just managing some calendars and inboxes and task lists and saving me some data entry. Not sure how that makes downsides gargantuan, or contributes to the internet dying. (Though obviously the bot will get worse as the internet continues to die if that's what it's using as a source)

> or contributes to the internet dying.

People will set these things to run wild on all platforms. Talking to real humans will be a luxury.

Eh… Titanic did flood in the engine rooms so… might work?

That humor aside: I think it’s about risk tolerance, and you configure accordingly.

You lock it down as much as you need to still do the things you want, and look for good outcomes, and shut it down if things get too risky.

You practice free love, but with protection. Probably still fun?

Big difference between running a bot with fairly narrow scopes inside a network available via secure chat that compounds its usefulness over time, and granting full admin with all your logins and a bank account. Lots of usefulness in the middle.

This is great news. The 1M context is much easier to work with than compacting all the time and seems to perform and remember quite well despite the insane amount of data.

The assumption with these weapons was that they would require too much energy to be portable enough to be undetectable in all of these circumstances (at least based on other reporting on the subject).

If the device doesn't require a lot of power, then it's entirely possible that American military commanders and research leadership would miss it.

Add to that an incentive to avoid helping the victims from a cost and overhead perspective, and you get a big ol' mess.

What I don't see in this article that should be explicit:

If your data is in this database, it's gone. Other people have it. Your sensitive data that you handed over to their teams has vanished in a puff of smoke. You should probably ask if your data was part of the leak.

Fail to see how a state actor would not have come across this already.

I anticipate they will fix this by adding better AI evaluation tools that work better to test their infra and changes.

In the meantime they will be quite a bit slower I’d imagine.

Also wonder if those seniors will ever get to actually do any engineering themselves now that they’re the bottleneck. :)

I loved HyperCard. Using it you really felt like you were building something of consequence - it was pretty magical to go from simple word processing to this in one model leap of a school computer.