Though AI is mentioned in other threads, there doesn’t seem to be one dedicated to discussing and sharing information on the topic. Whether one considers AI actually intelligent or not or “just” a method to distill vast amounts of data, some good, some bad, I do believe this technology is going to impact us all.
I have reluctantly moved from assuming it was just the “next big thing” in IT and would fizzle out as usual and become mainstream, to recognising that it is and will continue to have a significant effect across many diverse areas.
Some things AI-related are impressive. I use claude.ai to save me having to write code. You still need to know the principles, but it definitely saves you time.
Some things, less so. We have deployed our own AI instance at work, and asking it to create simple things like PowerPoint slides shows its weaknesses. I also managed to stump it with a simple Excel question recently, and it just got itself into a loop whereby it suggested the same two or three things repeatedly despite me telling it they didn’t work.
I think it shows that if you don’t want to use the public versions and can’t afford to spend a shit ton of money on it, then you get mediocre results.
However yesterday I saw a YouTube video about VEO 3 which really impressed
I felt that many aspects of his presentation were extrapolations detached from everyday reality. It reminded me of reading Alvin Toffler’s Future Shock about fifty years ago. The things that actually changed subsequent decades didn’t even exist then.
Several, rather than one, but we’re a heck of a lot closer than we used to be.
It’s debatable to what extent modern AI systems are truly “intelligent”, some problems seem to have been “solved” faster by an AI than humans but I don’t think there have been any genuinely paradigm shifting insights; they will probably come. Then we get into the whole thorny space of coming up with their own goals, morals, that sort of thing (Asimov’s rules are a good start, I wonder if the behind the scenes cutting edge research has taken notice of them - I bloody well hope so).
But, yeah, it’s worrying that we are on the cusp of Skynet becoming self aware.
“ For now, one thing is clear: how we interact with AI shapes how we interact with the world. Politeness isn’t just about getting better answers – it’s about reinforcing habits of clarity and respect in all our interactions.”
It’s about training you to defer to your new robotic overlords.
It worked in Japan - people there are still deferential because twitchy Samurai warlords were known for being inclined to remove peasants’ heads if they were not polite.
[warning, there mught be a completely unsubstantiated Internet myth embedded in this post but it illustrates the point]
This morning I was surprised to receive an unsolicited AI generated 5 minute long précis of one of my book chapters. The original chapter is available to download for free on the wonderful international academic research site, academia.edu. The précis was read out by one of academia.edu’s human employees.
Enjoyed listening and no complaints about the quality, but afterwards wasn’t surprised to see that they were trying to tempt me back to their premium service which tracks your mentions and downloads in much greater detail. I used it for many years, but as I’m no longer affiliated to a university, I don’t need that level of feedback.
Unfortunately, I don’t seem able to share the podcast unless I subscribe, but in the unlikely event of any one having an interest in the emergence of aesthetic landscape tourism in the Lake District, the original full written version is available on academia.edu at:-
Seems that in ‘creating’ a synthetic intelligence that can surpass our own abilities in so many ways, the crucial thing is defining the correct words to solve the control problem
“It’s important that we’re at least as good at defining a beneficial ultimate objective without unwanted consequences as we are at building superintelligent AI.
