Whatever, dude. Enjoy your protests and riots.

> I'm sort of surprised it took Jev for everyone to realize "oh yeah, this is a useful thing that we should have in our product portfolio".

Because it's especially parsimonious with token use.

> How do they plan on stopping me from using Claude code with the Claude in chrome plugin, or some other computer use agent?

Maybe they won't be able to stop you, but the average Joe won't be able to cover his tracks as effectively, or most likely, Claude Code and others will by default be required to assert the browser for the current session is being operated by an agent.

Average Joe will cover his tracks just as effectively. An average Joe will use the Claude Desktop app which opens a built-in browser where agent does the clicking - difficult to distinguish from a human doing the clicking.

yes, that's why I included this forward-looking statement.

> Claude Code and others will by default be required to assert the browser for the current session is being operated by an agent

> using a device that emits my TCP options

But what if you're connecting to a host using ssh keys that don't support all these extra options?

Is he known to talk in his sleep?

operating coding agents on the primary machines of finance professionals runs directly into the lethal trifecta.

I've professionally interacted with GS for over 30 years. It blows my mind that GS allowed real exposure to "the lethal trifecta" for even a moment in time.

> In spite of all of these existing issues, it seems like Anthropic has worked at a feverish pace to pack more features into Claude Code without much care for quality or stability

To me this is why AI-pilled orgs or groups won't win.

> Foot down, get to 10 trillion parameters, see if the scaling holds, if so off to 100 trillion parameters. If the scaling continues to hold we'll be asking the AI to do the exploratory research for us soon enough.

Where's the money coming for those 10X and 100X ventures?

From mass layoffs of tech company employees.

There’s a good example of a model that won’t scale.

It only worked for Milken because he had so much money as to be bulletproof and could buy his reputation back.

The way I look at AI R&D is it's a really clever brute force search of the problem and solution space. The amazing bit to me, is how rarely, these models and methods seems to get stuck in local maxima. i.e. ex ante, I'd expect these models and methods to get stuck in local maxima 100% of the time.