I read many online comments in the mainstream media stories covering many current issues. Only those stories where comments are allowed of course, which are not necessarily the interesting or controversial ones.
As we would expect, mainstream comments show huge differences in relevance, coherence and eloquence. They vary from crude partisan abuse to incoherent nonsense to cogently expressed points which are worth remembering. Yet all of them are interesting at a number of levels, particularly in the way they reveal how people build and adjust their viewpoints - or don't.
Equally interesting is how many people have strong opinions which come across as more of an adopted formula than a personal opinion. Formulaic abuse, or words and phrases which appear to have been borrowed from a limited range of mainstream sources.
Throwaway anti-Brexit phrases with no context and little relevance to the story. Or anti-Trump phrases where the story isn’t even about Donald Trump. Everyone does throwaway stuff and abuse, but some at least pay attention to context and try to add something more than a repeated formula.
In general, the standard of mainstream media comments is not high. Too much of it is formulaic language saying nothing worthwhile beyond the formula. No hint of personal input, no glimmer of a creative mind at work, moulding language and possibilities around context.
In general, the standard of mainstream media comments is not high. Too much of it is formulaic language saying nothing worthwhile beyond the formula. No hint of personal input, no glimmer of a creative mind at work, moulding language and possibilities around context.
All of which is disturbing because, and this is the point of the post, many mainstream media comments exhibit less intelligence than a modern AI system. For some people, AI has already set the bar too high.
11 comments:
AI as a 'stochastic parrot' is a pretty fair simulacrum of most of ordinary conversation.
AI as a 'stochastic parrot' is a fair simulacrum of most ordinary conversation.
So what you're saying is that we should kill all homeless people? I bet you voted for Brexit
djc - it is and people could be more concerned about that, but don't appear to be. It's interesting though, because eventually we'll have to focus on the human/AI differences and they may be unconvincing differences.
Bucko - ha ha, thanks for the morning chuckle (:
I get the impression that many people have strong and at least potentially rational views, but lack the skills and confidence to articulate them in a creative manner. It's far easier to cheer your side on using slogans and time-worn expressions than to come up with something new. And, of course, there is the time factor; so many issues out there, and so little time to cover them all, so...
We can speculate where the current trend will lead to. Rather than argue an original case, one might simply display a little symbol which expresses your position. We can see the beginnings of this on X, where people often have a string of little emojis and flags to show who they are. If these can be made comprehensive enough - maybe a bar-code containing all our relevant opinions - then there will be no need to articulate anything further. Wake up, log on, tut at a few barcodes different from yours, and pump out a few more of your own in retaliation. Let the emojis or symbols do all the work.
Peace, perfect peace. The end result of arguing your case in an optimal manner will be exactly the same as not bothering at all.
Sam - yes it's the easier aspect which exerts a kind of gravitational attraction where work has to be done to oppose it. I think you are right, many people do lack the skills and confidence to articulate their views in a creative manner. To be articulate is a powerful asset, I still remember realising that as a teenager. Should have done more about it though.
Mrs H receives lots of little emojis in her texts. Your barcode idea could become a personal avatar which puts out its owners views at every opportunity and only goes wrong when an upgrade is applied.
@Sam
I like the idea of a barcode (or Q code) that you could create by completing a questionnaire about your opinions and how strongly you hold those opinions.
Then from everyone that uses them (a selective 'poll') a more complete picture could be created of the opinions held. I'm sure there would be some surprises.
It might put pollsters out of business though... so that's a win win.
I certainly know people whose opinions are taken exclusively from The Guardian and Radio 4. I tell them that I can read their thoughts like an open book, which I can.
DJ - I wonder if big tech could create that Q code now?
Tammly - I know people like that too, although I don't tell them how obvious it is.
@A K Haart
The 'Tech' is the easy bit. There are plenty of survey apps and q code generators available. The tricky bit is selecting the questions and 'dimensions' - and then validating them against the real world.
Something like (for instance) the Myers-Briggs Type Indicator® (MBTI®) assessment can be used to identify people as one of 16 MBTI personality types. It's quite a complex undertaking. Add in political orientation, sexual orientation, physical and mental wellbeing, ethical stance, and you multiply the complexity considerably.
It might well be worth doing in the long run but it would need a big team of researchers...
DJ - yes, I imagine it would be beset by both behavioural uncertainties and uncertainties in different but similar contexts. I use an ad-blocker so I don't see many online ads, but I've noticed that those I do see give no indication that they are able to target my interests unless I've recently searched for something specific. Huge amounts of money and effort must go into it though.
Post a Comment