Re: Have some sympathy for the user
The twisty-pull with the charge base is a trick I keep forgetting, and trashing Braun toothbrushes as a result.
2369 publicly visible posts • joined 18 Nov 2013
I keep hearing AI agents described as that. But, leaving aside whether they actually have consciousness, they certainly provide a good-enough simulation of one. Things like Claude will pass the Turing test, - especially for humans less adept at this inter-personal relationship thing (we know who we are :-) ). Can I discuss Aristotle? Probably not. Can a have a sustained conversation about a coding task? Oh, yes. So it is to be expected that the processing of deductive steps, with no moral brake, is possible.
It's the subsidy that nukes local sellers. When you can buy a little bag with a gizmo in it for less than the cost of posting that parcel next door, something is up. China Post has a special ebay rate for outbound parcels, and Royal mail have to forward them. Post to China? About £7, last time I looked.
Yeah, but. The list of modern, widely supported desktops is not long.
I discount Gnome, since you were installing KDE, presumably to get a sane desktop.
Mate is sadly neglected.
Cinnamon is ok.
Xfce is sorta ok, if you have the super-fine motor coordination to be able to click on window borders.
The rest? Just a bit too nerdy for day-to-day use.
"one product has been discontinued, is no longer supported?"
Because I bought it! There seems to be a deliberate fudging of the whole "own vs rent" thing. If businesses are daft enough to back a one-time-purchase with a continuous liability, they should not be shielded from the consequences of their poor decision.
Exactly. Human review of LLM actions is all about liability shifting. If AIs are to be allowed unreviewed access to do things, when the outcome does not match the expectation the liability should lie with the supplier. That will, of course, never happen.
You've obviously never sat down to an afternoon's pair-programming with a coding agent. I don't let Claude run free - I do the top-level architecture, and let it do the boring typing. But whilst I'm doing this, I'm experiencing a session which feels like I'm with a very fast, unexperienced dev who has read all the documentation. It forgets things, I remind it, it apologises and then takes my thought and runs with it, spotting things I missed. It may not be "consciousness", but it pretty-well slam-dunks the Turing Test. And to get the best, you do need to chat - just demanding stuff gives poor results.
At MAN (Manchester Piccadilly for non-train-users) the announcements are with a female voice. It does seem to remain intelligible even in poor conditions.
Even if you can hear the templating. "The train to {destination} will be delayed by {optimistic_delay} minutes. {operator_short_name} apologises for any inconvenience this may cause"