How do you think AI will change things?

[Mountainhowl]

Not an AIBU just the best board I've found for traffic and discussion :)

I'm curious to know where others think AI will take us? How will things change and how do you think governments will deal with the changes?

Will there be mass job losses? Do you think we will get a UBI? What jobs do you think are likely to be safer?

Will life get better or worse?

Do you use any form of AI? I occasionally utilise chat gtp (the free one) for brainstorming ideas and help writing things like blogs for my small business, and I use some of the AI functions available in Photoshop/lightroom for image editing

We'll get loads more new music from artists who are dead. Just look for ai covers of Freddie Mercury or Frank Sinatra on YouTube, those sort of things will get way better and more widespread.

I’ve read a couple of immigration threads today and they seem devoid of any AI mention

Talk of need for workers and falling birth rate but surely the missing piece of the puzzle is AI

Need for workers will drop a fair bit

I'm not talking about now, more of a possibility for its use, how things might change etc.

One thing you can say for sure is that everything we predict will be wrong. Humans are staggeringly bad at predicting the future. Awful. We seem to get everything wrong.

Us worker drones will get ripped off even more efficiently by the rich and powerful.

marking because I am really worried about this. Need to educate myself.

AI has already started to impact the job market and the economy. While some jobs may be lost, new ones will be created. According to a report by McKinsey Global Institute, AI has the potential to deliver additional global economic activity of around $13 trillion in the foreseeable future and by 2030, or about 16% higher cumulative GDP compared with today1. However, the same report also states that over half of LinkedIn members worldwide are likely to experience job disruption or augmentation due to AI2.
Governments are already taking steps to regulate AI. The UK government, for example, has proposed a pro-innovation approach to AI regulation3. AI can also help policymakers make more informed decisions4.
Universal Basic Income (UBI) is a concept that has been discussed as a possible solution to job losses caused by AI. Some experts believe that AI will create so much wealth that every adult in the United States could be paid $13,500 per year from its windfall as soon as 10 years from now5. However, the implementation of UBI is still a topic of debate and requires further research.
As for which jobs are likely to be safer, it is difficult to predict. However, AI can be used to enhance workplace safety by monitoring worker fatigue, detecting potential hazards, and keeping workers safe at all times6.

😂😂😂 accurate

Thanks for this loads of info

One main concern is we work out how to capture the windfall here

I can’t see how societies will function if they don’t get funds to off set job losses

Don't thank me, I didn't even read it.

I pasted the question into Bing and that's what its AI came up with. 😂

Tbf I won’t either 😂

But I was impressed with the effort. I was going to ask if you did it all

The general point is a good one - we need to get money from AI here

Students cheating on every assignment is problematic...

I think that may be outdated thinking. Perhaps we will value the skill of researching things more than the ability to commit them to memory.

Surely cheating on assignments will be the least of our problems?

?

The future can only come from the present. Unless my understanding of spacetime is off.

And right now, AI is shite. And shows no signs of actually being able to shake off all the artefacts of intelligence that humans show.

I haven't heard, read, or spoken to anyone who thinks they have a clue about the problem of stopping the "intelligent" part of AI following the path that natural intelligence has taken. Only (a) a lot faster and (b) more efficiently.

We're still at the stage where asking an AI model to suggest ways of reducing domestic accidents returns the suggestion of removing the top and bottom stairs.

@Lentilweaver well smartphones have certainly dumbed people down. I dread to think what happens when people barely need to engage their brains and just start regurgitating info provided by aI.

I saw that this little puzzle throws up nonsense from aI. I tried it myself and it can't seem to work it out at all.

I have 6 eggs, I broke 2, fried 2 and ate 2. How many eggs do I have left?

I'm confident that aI isn't terrifying yet.

Those saying that AI isn't that clever yet, what versions are you using? Chat gtp4 is supposedly around as smart as Einstein, and assuming the technological growth of doubling every 18m or so they expect gtp 5 to be twice as smart as this. The free version is gtp3.5, Im not in the position to pay to use gtp4 so haven't tried it myself

(by 'they' I mean Mo Gawdat, he's easy to listen to and I've just been watching another interview/podcast thing, I've seen others saying similar however)

At moment I mostly encounter AI when it's trying to fake being human (fake survey respondents trying to get thank you vouchers).

And fake commentaries and scientific article reviews. Which have extremely low usefulness.

and fake social media accounts ( trying to say angry things to make the real humans angry). So that's 3 instances where AI is being used to perpetuate fraud. I wouldn't mind AI-peer-reviewers if they actually said something useful, btw. But useless & fraud is annoying AF.

I know "machine learning" is being used for pattern recognition in a lot of sciences, applying probability rules and processing lots of data quickly. That seems like good thing, exactly what AI should do well.

AI is good for translating between some languages & sometimes composing 1 sentence email replies.

I can't think beyond those examples. Since most of my experience is people using AI for fraud, I'm not a fan at moment. We shall see.

AI could really help in education. When I mark maths books and exam papers, I can’t remember exactly how well every student does on every topic. I can record lots of data, but then the amount of time I spend doing that comes off planning time. I can see AI both marking work and tracking progress, showing areas of weakness, keeping track of which topics are fading in students memory. Particularly in the run up to exams, they could really tailor the revision process. We already have programs that do this, but they’re both clunky to enter the data, and then also create too much data to effectively analyse and utilise.
I’m not sure AI would be effective at behaviour management, or how well it would be able to adjust explanations when they weren’t working with a particular student. I’d love to think AI could be made to make education better, rather than used as a cheap alternative to human labour. I think you could probably make a pretty soulless online maths tutor that would do an okay job of teaching maths to reasonably compliant students. A well trained teacher using AI to make their lessons crafted perfectly for each class would be a massive upgrade to the tools we currently use.
I’m just imagining something like getting the AI to create seating plans for a specific lesson, based on groupings by behaviour, Sen and medical requirements, and ability/common misconceptions in the topic being covered that lesson, then selecting a selection of questions that work for each group, along with all the solutions.

We’re not quite there yet. AI hasn’t quite grasped the concept of things being right or wrong and a lot of students work looks like the badly written words we use to prove we’re not robots! I don’t think we’rea million miles away, though.

How does it work with spreadsheets? Could you point me in the right direction eg links?

AI is nowhere near as clever as Einstein. There’s a massive difference between answering questions about relativity and coming up with the theory in the first place. If AI was as clever as Einstein, e we would be hearing about all the scientific breakthroughs it had come up with. Currently, it just regurgitates pre-existing information in new combinations. It isn’t yet capable of truly imaginative thought. It can make predictions and draw conclusions based on data you feed it. It can’t decide to analyse it using a new measurement of correlation it’s invented itself, for example, but it can spot connections that we’ve never thought to look for.

Mind you, I haven’t done much reading up on AI for about three months, so I could be totally out of date by now!

I Use ChatGPT 4 subscription.

It is invaluable for doing grunt work - parsing, formatting, writing code from examples - all the stuff they waste years at school teaching. You still have to know how to ask, what the code you ask for does and know if it's right. Otherwise it will simply .....make...shit...up.

Half my industry is training AIs to write reports. The other half is training AIs to read reports and summarise or answer questions about the content.

Eventually they will put two and two together 😂

I work in digital transformation in healthcare. AI is incredibly exciting!!

Imagine having AI that can read someone’s genome to detect a persons risk factors, create the perfect pill for them, save doctors (and patient’s time) by looking at scans. The potential it has blows my mind.

But for now at work we are just using robots for data transfer in EPRs and ChatGPT to answer our silly questions we can’t be bothered to google haha.

If anyone has time this is worth watching, the interviewer is Steven Bartlett from Dragons Den.