Re: "At present, Windows Recall feels like it was put together with insufficient thought"
I fear they've been entirely cleansed of any trace of ethics.
42421 publicly visible posts • joined 16 Jun 2014
You can imagine some techy getting excited with his idea that "We can already take screenshots when the user asks. Why not do it automatically and just save the results?" Or maybe just doing it and showing the higher ups. The question is why there weren't any grown-ups in the room to realise it was a Really Bad Idea?
The nature of the offer seems to be compensation so it appears to be a civil case. A guilty verdict would only apply in a criminal case. Even if a guilty verdict were an option then, as they're wanting it to be tried by a judge as opposed to not tried at all it wouldn't make a difference. Very likely the motivation is that by offering to concede the case before a judge they can avoid having the details of the case aired in public.
"
The risk here is that first you have to prove something is an AI."
Take the vendor's word for it. Having made a big fuss about their wonderful AI they'll be in no place to claim the legislation doesn't apply. Conversely the need to make the big fuss will ensure they do just that.
OTOH I, like many others, maintain a web page which is a programme of planned events, in this case a local Civic Society's talks programme. If we're lucky* we start off each September with a full programme up to May. Every month the list gets a bit shorter and the poster image changes. There's no point in keeping that page unchanged in perpetuity; if you want check up on last month's talk it's too late already.
Some stuff really is ephemeral.
" Since Covid finding suitable speakers has become a bit more fraught than it used to be.
So who is to blame for that fact that locally the said telco is putting its finances into rolling out FTTP to the premises well served by FTTC but customers on the further reaches of the same cabinets are struggling with poor quality service?
Even accepting the reliability of limitations of locally powered digital over POTS this should not have been considered in advance of getting an adequate digital service to every location. Not just the easy ones, all of them.
OK, let's pursue that idea a little.
You're going to require a jury for the trial. These days potential jury members are going to be drawn from a wider spread of the population and attitudes than when death sentences were normal, at least in societies like the UK. The prosecution then has the problem of risking one member of the jury who, whether from principle, awareness of the possibility of miscarriage of justice* or squeamishness, is not prepared to vote for killing someone and getting a not-guilty verdict against even the strongest possible evidence.
So how are you going to deal with that? Allow for majority verdicts? Given the risk of miscarriage of justice for a capital offence public opinion might not allow for that. But if majority verdicts are to be banned for one category of offence perhaps they should be banned altogether? You'll open up a can of worms going down that route.
* On a personal note I found it stressful enough doing my job investigating crime whilst being aware of the possible consequences of an error. Doing it under the shadow of a possible capital offence would have been impossible.
"The fact that online fraud in the UK doubled year on year in 2023 to a value of £2.3bn shows that there's more cyber crims and they're mostly getting away with it."
This is not surprising.
Take, for instance dodgy phone calls. There is no single reporting mechanism. The subscriber is supposed to triage the call themselves and then decide which of the reporting sites is appropriate. Receive and attempted fraudulent call? AFAICS there is no site for reporting this. If you have actually been defrauded there's a site to report it. Not collecting details of attempted frauds tells me there is no attempt to collect intelligence that might enable fraud operations to be detected and closed down sooner.
Take, as another instance, clickable links in spam. Are the public being discouraged from clicking them? No some financial institutions are routinely sending unsolicited emails with valuable marketing information spam to customers, training them to believe that a link in an unsolicited and unexpected email that appears to be from that institution can be safely clicked.
Are TPTB discouraging this? No, they're at it as well. IME a visit to any NHS service will be followed up by a text* with a link for feedback. And let's not forget sending texts to a landline where the text can be picked up by whoever's nearest the phone but their interpretation of GDPR prevents them, when queried, from saying who it was intended for.
* There's a great deal else wrong with this. The text doesn't say who it's intended for and just says "your recent visit" so anyone wanting to fake it can just spam them out blindly. Even for its intended purpose this fails if there have been a few appointments in quick succession before the first text arrives. Also any hospital appointment will ask for next of kin contact details and my local trust would (I think I have now dissuaded them) treat this as the contact information for the patient.
"Would you give up some hacker you knew to Xi Ping ?"
You're thinking as a law-abiding person, not as a criminal to whom the $10m is very likely aimed.
Several possibilities.
First, just considering Russian criminals: Criminals don't really stick together and some would easily be tempted by $10m if he could be kidnapped and transported across the border. There might even be the possibility of eliminating a rival or getting their own back if they've been cheated.
Secondly the $10m might be enough for a non-Russian gang to attempt the same.
Thirdly, there's the possibility of a sting operation, say somebody in a neighbouring country looking for help to st up a new operation and offering to cut him in. There are one or two countries where he might feel safe to visit.
Arrest, trial and lengthy sentences would be what I'd term a "more aggressive" method of takedown.
20 years of supervised release doesn't seem as much of a deterrent as 20 years imprisonment although I suspect there might have been quite a bit of trading to get there.
And with a $10m reward I wouldn't rule out someone ordinarily resident in Russian suddenly turning up somewhere more accessible to arrest.