Re: tendency to generate false information – a phenomenon known as "hallucination."
Right now AI is the term being used to push these things onto the world in general. Yes' it's an oxymoron but that's the way marketing works.
43025 publicly visible posts • joined 16 Jun 2014
A statement based on reports that somebody reported a crime or was a witness at a trial and mangles them to say that that person was convicted of the crime is simply not true. Which epithet you choose to apply to the untruth is your personal choice and frankly doesn't matter very much. What matters is that people are trusting the system that creates such a statement.
"How much of a world model or theory of the world does a newly conceived human have?"
The newly conceived born human acquires a world model by being a physical entity and encountering the other physical entities around it. It starts to do so before starting to acquire language. Our non-human relatives do the same without ever acquiring language. Language - words - is/are the means by which we apply symbolic labels to the real world in order to build and manipulate ideas about them. They are not intrinsic to an internal model of the external world. But words are al LLMs have. They do not have the physicality to react with the physical world.
“take full control of our device” sounds more like Microsoft's approach. I think Apple is more along the lines of "it's your device but you're not going anywhere outside our walled garden". I suppose vulnerabilities like these could have also opened up the possibility of 3rd party stores such as F-Droid run for Android phones.
People have been using textual analysis for years to (try to) identify writers. Computers have been used in the Who Really Wrote Shakespeare game since they became available. But I don't see it working well by trying to score individual sentences. I'd have thought it would work better picking on a whole lot of factors, consistency of choice of grammatical constructions, consistency of shortening (e.g can't vs cannot vs can not) & so forth that would require gathering statistics from the entire text or at least substantial portions of it. It might be quite reasonable for all replies form any particular LLM to occupy quite a small portion of a multidimensional space defined in that way.
An over-seas allowance? I'd have appreciated that. Instead I was up against the Civil Service/HMG mentality which says "We rate people who are qualified for their jobs by their education well below those who aren't, promotion comes with responsibility measured by direct reports, not by what the actual job is, and if we want a national pay policy the easiest place to start is with the public sector."
Scene of crime visits could be nasty and some of the nasty bits got brought into the lab. A colleague had a sofa brought in; it stood around for weeks for some reason and had to be given a wide berth - the little buggers can jump. In fact fleas were a (fortunately infrequent) occupational hazard. But the nastiest job was removing bits of a bomber who'd scored an own goal from bits of his clothing so the latter could be shown to his family for identification.
Amazon is going to show you what *they* want you to see and buy,
They may succeed with the first but fail with the second. Perhaps they could try an A/B test with an alternative search engine that does as it's asked to find out how much trade they're turning away.
"it's probably best just not to answer the call"
No, siita has the right idea. Get them to hang on. While they're doing that they're not bothering anyone else so you can put it down as a public service on your part, and it's wasting their time. Given how few such calls we get my belief is that there are lists circulating of numbers it's best not to ring.
I keep saying this but:
The telco has records of at least who it took the call from. What's needed is a number, say 1476, to ring immediately after hanging up to report the spam. The telco records the number and with a bit of statistical checking to see if the source is being reported enough times it credits the reporters' accounts with a fee for their troubles (at least double for numbers on TPS), adds a fee for its own troubles and bills whoever sent it the call. If it's another telco then they can add their own fee and pass the bill back to the source until it either reaches the caller's account or a telco which has not yet learned to keep records. This will make it unduly expensive either for callers or for telcos who are prepared to handle the calls carelessly. One way or another it would kill the whole business stone dead.
In practice it probably wouldn't need to be implemented. It would put telcos to considerable upfront costs to set up the system. Once legislators or regulators started talking seriously about implementing it I think the industry would quickly discover other, effective ways of stopping such calls, ways which up to now haven't been practical.
"although I never answer calls from numbers I don't recognise"
What you should do is answer ask them to hold on - you need to look something up to answer their question, have to go to answer the door, whatever - and then put the phone aside for 10 minutes. AFAICS that gets your number on an even more valuable list, the list of numbers it's best not to call. It certainly works for me as we get very few.