Editorial – June 2022: Truth And Untruth: Some Of Which Is Outright Lying.
Hello everyone
There are times when I wish one of AE Van Vogt’s predictions for lie detectors to be commonplace in his ‘Null-A’ books had come true. Although they were there to verify honesty and specifically demonstrated lead character Gilbert Gosseyn’s identity and failed, you would have to wonder what was the turning point for them to have been introduced in the first place. Obviously, people had to be lying a lot to necessitate such a device to be commonplace. Even so, the implication is that lying is a normal state of affairs and the use of a lie detector when someone is called out to be a liar. I should point out that these lie detectors aren’t anything like the polygraphs we have today. In Van Vogt’s future, a hand touch and AI interpretation no doubt read more than pulse and body temperature but a brain scan looking for ‘disruptive neural flow’. That was back in 1948. Not too crazy as we do know now how the brain reacts differently when telling a lie. Most people pause a few split-seconds when telling a lie and maintain the falsehood.
These days, the detection of a lie is comparing to known truths and mismatches to outright lying is complicated. There are a lot of lies for all sorts of reasons and not necessarily in self-interest. On the Internet, this isn’t helped when there is so much of a mixture of truths and lies. Consequently, you need good judgement to be able to tell the difference or have access to a lot of things off-line as a means to double-check and then look again. Trust is an odd word on-line, especially when so many tend to think of other people as being akin to themselves, not thinking of the con artists who will use that against you. Honing the ability to tell the difference can often be painful and expensive and a stronger reminder to keep up-to-date with the latest con tricks that are employed.
Even so, there is a need for better honesty on-line but even the AI algorithms have problems or when you do a google picture scan, there’s a large percentage of pictures that are clearly not the subject. The AIs need to improve a lot or at least go by more than a file name in identification or a name mentioned in the text. How can they determine fact accuracy without knowing what is true or false will be a stumbling block unless there is some human agency for it to confer with or a reliable source. Said human agency would have to be honest and probably need more than one of them to ensure there is no hidden agendas. All of them adds up to a lot of time and that would just be covering current events.
Then there is the problem of overcoming a wrong belief. Discovering recently that 2% of the American population believes that the Earth is flat and the sun rotates our planet makes me wonder how did said people slip pass the education net? They certainly can’t be on the Internet. After all, the world is on the back of giant elephants or, at least, that is what some others believe. Could an AI convince these people any differently where even humans have failed? Physical evidence would be one thing. Walk one of these people to the horizon doesn’t prove the world is round because the horizon is rarely if ever reached. To them, it would mean that the world is much bigger than they thought. Taking them up the highest mountain and hope there is no cloud would just indicate that they can’t see far enough. Could I convince them to go up in an aeroplane or rocket so they could see the Earth at a distance? They’d probably think it was faked. The world, not the plane, unless they were prepared to sky jump. Parachute not optional. A simple acid test that both human, all right, nearly human and AI would fail to convince a person against such a belief. If nothing else, it does set a limit on how to convince some people of a simple truth before getting onto something more complicated.
This made me do some deep thinking as well as checking on-line where the Moon is used as evidence that if it is spherical than so it the Earth would follow the same lines as the Sun is also that shape. The same would also apply to any of the planets seen through a telescope. Then you would be hit by just because they are, doesn’t mean the Earth is as well. Then again, you would think space flight would be fool-proof but, undoubtedly, these are people who also don’t have much scientific knowledge. A country to spend that many billions of dollars over several decades to sustain a lie has to be preposterous let alone so many people not letting the truth out. For mobile phones to function, they rely on satellites which is one of the results of space technology. A false belief can be sustained from what they see as evidence and a few people supporting it versus, well, everyone else. The way Internet AIs can segregate you with your apparent ‘tastes’ on-line could potentially do even more damage unless you’re prepared to ignore its choices and just explore and learn. If you don’t believe that, there are some American Net Servers that don’t allow their clients to see anything of the rest of the world or at least only what they want to show them. An Internet equivalent some of the American TV news channels if you like.
Under normal circumstances, the above probably wouldn’t be used as a test to distinguish between truth and lies. Even so, you should be able to see the problem of programming an AI to tell the difference in the first place and be able to explain it like I did above. No magic involved. It raises a question that no matter how powerful you made an AI, would it be possible to tell the difference itself. No wonder the Internet is going to need some means of proving the difference.
If anything, having an AI being able to differentiate between truth and lies is going to be even harder than a human being able to tell the difference. Oddly, people have shown themselves more prone to trust an AI than a human, forgetting the former is programmed by the latter. It does raise interesting questions as to how easy it would be to con an independent AI. I suspect on some things, like the Earth being globe-like and other scientific facts probably not because there is plenty of verification. Anything else would depend on building up suitable proven undeniable database evidence and a screen flash of approved information when you look something up, although I think this would also be backed by percentage accuracy. There would also be a need to build up evidence and that would also mean being provided with the right algorithms and not take anything at face value. If anything, it might be easier to tell an AI not to take any human on trust. Indeed, that would also apply to humans to each other. At least it couldn’t be talked into giving money in a digital friendship or likely to send money to a Kenyan bank account.
If nothing else, we need a lot more work on the Internet to differentiate between good and falsehoods and raise question marks on unverified info. Any such AI shouldn’t dictate what you want to believe in but should provide a counter-argument on some subjects for verification. Would it turn people towards becoming flat-earthers? No, at least not for Net-users. We know better than that but knowing the odds of something being correct or wrong is likely to have a better effect on future generations. Either that or we’re going to give rise to a generation of really ignorant people and do we really need that?
Now, if something can come up with something that will sort out each person can only have a single identity on-line or alias verification then there would be a level of improved trust.
Thank you, take care, good night and try to be honest as that would surely confuse any AI observation.
Geoff Willmetts
editor: www.SFCrowsnest.info
A Zen thought: Life needs a world without bullets.
What Qualities Does A Geek Have: Extreme observation.
The Reveal: I know, I haven’t mentioned Putin’s sanity at the stop of the editorial this time. It’s not forgotten nor do I think the longer he doesn’t initiate a third nuclear war that he might do it. A hooded cobra waits to strike when you least expect it. So this time, I’m putting my own world comment where I least expect it.
Observation: People think Darth Vader is from German when it is actually dutch. I hadn’t realised that ‘Vader’ was Dutch for ‘father’ and ‘Darth’ means ‘dark’. The Dutch have multiple choices for ‘light’ so there is probably no equivalent.
Observation: In the UK, Channel 82 shows a lot of old UK films. Keep an eye out for a 1978 kiddie 54 minute film called ‘A Hitch In Time’. Apart from being the last film actor Patrick Troughton did before he time and was a time machine inventor, you need to pay attention to the computer console. It came originally from ‘UFO’s SHADO’s Moonbase, usually in front of Joan Harrington’s desk and, yes, the lights did come on, especially the zig-zag lights as shown in its opening credits.
Observation: Looking at the Bela Lugosi Dracula and his successors with what we know today, vampires tend to wear what they were wearing when they crossed over, presumably because it has the same durability as their own bodies and would come back when resurrected. Considering Dracula was wearing an evening suit would suggest he was off to an evening meal or even the theatre when he was bitten. Make a note that should you be turned into a vampire, don’t submit when in a bath or shower. You might well be the first streaker vampire.
Will Make You Think: Can androids be put in the washing machine under ‘synthetics’?
Computer Observation: Did you know there is a way to create macros in Windows 11. In the ‘Search Programs And Files’ slot type in ‘Power Automate’ and you will have to wait for it to check and put in updates. It will then run and ask for your email address and confirm your password and ask if you want it to save info for Microsoft. It doesn’t send what you do just an operational check. The only problem is thinking up a way to do something effective with it. I’m not even sure if it can replicate certain functions of moving files into different directories without knowing which ones.
Feeling Stressed: Treat an answer or more to one problem at a time.
