Comment by notepad0x90
1 day ago
ML has been used for influence for like a decade now right? my understanding was that mining data to track people, as well as influencing them for ends like their ad-engagement are things that are somewhat mature already. I'm sure LLMs would be a boost, and they've been around with wide usage for at least 3 years now.
My concern isn't so much people being influenced on a whim, but people's beliefs and views being carefully curated and shaped since childhood. iPad kids have me scared for the future.
Quite right. "Grok/Alexa, is this true?" being an authority figure makes it so much easier.
Much as everyone drags Trump for repeating the last thing he heard as fact, it's a turbocharged version of something lots of humans do, which is to glom onto the first thing they're told about a thing and get oddly emotional about it when later challenged. (Armchair neuroscience moment: perhaps Trump just has less object permanence so everything always seems new to him!)
Look at the (partly humorous, but partly not) outcry over Pluto being a planet for a big example.
I'm very much not immune to it - it feels distinctly uncomfortable to be told that something you thought to be true for a long time is, in fact, false. Especially when there's an element of "I know better than you" or "not many people know this".
As an example, I remember being told by a teacher that fluorescent lighting was highly efficient (true enough, at the time), but that turning one on used several hours' lighting worth of energy for to the starter. I carried that proudly with me for far too long and told my parents that we shouldn't turn off the garage lighting when we left it for a bit. When someone with enough buttons told me that was bollocks and to think about it, I remember it specifically bring internally quite huffy until I did, and realised that a dinky plastic starter and the tube wouldn't be able to dissipate, say 80Wh (2 hours for a 40W tube) in about a second at a power of over 250kW.¹
It's a silly example, but I think that if you can get a fact planted in a brain early enough, especially before enough critical thinking or experience exist to question it, the time it spends lodged there makes it surprisingly hard and uncomfortable to shift later. Especially if it's something that can't be disproven by simply thinking about it.
Systems that allow that process to be automated are potentially incredibly dangerous. At least mass media manipulation requires actual people to conduct it. Fiddling some weights is almost free in comparison, and you can deliver that output to only certain people, and in private.
1: A less innocent one the actually can have policy effects: a lot of people have also internalised and defend to the death a similar "fact" that the embedded carbon in a wind turbine takes decades or centuries to repay, when if fact it's on the order of a year. But to change this requires either a source so trusted that it can uproot the idea entirely and replace it, or you have to get into the relative carbon costs of steel and fibreglass and copper windings and magnets and the amount of each in a wind turbine and so on and on. Thousands of times more effort than when it was first related to them as a fact.
> Look at the (partly humorous, but partly not) outcry over Pluto being a planet for a big example.
Wasn't that a change of definition of what is a planet when Eris was discovered? You could argue both should be called planets.
Pretty much. If Pluto is a planet, then there are potentially thousands of objects that could be discovered over time that would then also be planets, plus updated models over the last century of the gravitational effects of, say, Ceres and Pluto, that showed that neither were capable of "dominating" their orbits for some sense of the word. So we (or the IAU, rather) couldn't maintain "there are nine planets" as a fact either way without grandfathering Pluto into the nine arbitrarily due to some kind of planetaceous vibes.
But the point is that millions of people were suddenly told that their long-held fact "the are nine planets, Pluto is one" was now wrong (per IAU definitions at least). And the reaction for many wasn't "huh, cool, maybe thousands you say?" it was quite vocal outrage. Much of which was humourously played up for laughs and likes, I know, but some people really did seem to take it personally.
9 replies →
I think the problem is we'd then have to include a high number of other objects further than Pluto and Eris, so it makes more sense to change the definition in a way 'planet' is a bit more exclusive.
Time to bring up a pet peeve of mine: we should change the definition of a moon. It's not right to call a 1km-wide rock orbiting millions of miles from Jupiter a moon.
1 reply →
[dead]