AI: Good or bad? (1 Viewer)

[Brighton Sky Blue]

Just watching some videos on YouTube and noticed there's now AI summaries provided if you can't be arsed to watch it. On balance will AI do more harm or good? Besides the carbon footprint of powering it there's the problems of it making people even lazier than they are now. Equally it can be quite useful for taking time consuming but unchallenging work out of people's hands so they can focus on other things.

Not sure overall, what do folks think?

 

---

[Mcbean]

AI has completely fecked Facebook as people just load crap

 

[Sick Boy]

It's already too widespread and not used in the way it should be. In its current form, it will do more harm than good.

 

[Brighton Sky Blue]

It's already too widespread and not used in the way it should be. In its current form, it will do more harm than good.

Noticed that some recruiters are just behind the curve on people filling in job applications with it.

 

[Sky Blue Pete]

Good

 

[covcity4life]

Well I love it as a teacher

And I use it for gym and diet stats etc too

 

[Marty]

Never used it as far as I'm aware other than summaries on googles search function. Some of them are already getting very realistic. As with everything, it's how it's used. For the current totalitarian government we currently have, they'll use it infringing on peoples lives.

 

[Brighton Sky Blue]

Well I love it as a teacher

And I use it for gym and diet stats etc too

I used it for report writing, wrote the same quality that I'd normally do in a fraction of the time.

 

[Sky Blue Pete]

Write sermons using it so so helpful
Also it’s better to refer to it as automated machine learning nothing artificial about it

 

[Joe King]

Hate it.

 

[Otis]

I do use it, much to the chagrin of my daughter and my missus' daughter. They absolutely detest it.

I do feel guilty using it. It can be quite damaging for creatives.

I know a guy who does artwork. Album covers, band name logos etc. AI can conjure pretty much the identical thing in a fraction of the time and for free.

I have been using it for theatre backdrops.

There's quite a furore over Tilly Norwood at the moment. An AI actor.

Tilly Norwood - Wikipedia

en.wikipedia.org

 

[Brighton Sky Blue]

Write sermons using it so so helpful
Also it’s better to refer to it as automated machine learning nothing artificial about it

Didn't know you delivered sermons? Fair dos

 

[Brighton Sky Blue]

I do use it, much to the chagrin of my daughter and my missus' daughter. They absolutely detest it.

I do feel guilty using it. It can be quite damaging for creatives.

I know a guy who does artwork. Album covers, band name logos etc. AI can conjure pretty much the identical thing in a fraction of the time and for free.

I have been using it for theatre backdrops.

There's quite a furore over Tilly Norwood at the moment. An AI actor.

Tilly Norwood - Wikipedia

en.wikipedia.org

This is the problem isn't it. Humans need a creative outlet and this risks taking it away.

 

[fatso]

Potentially it could be very good, but humans being as they are, some will use it for bad purposes.
The effect on the job market could be disastrous if it replaces too many jobs too quickly, so it remains to be seen if its helpful or not.

 

[Captain Dart]

It has no creativity as it uses mainstream thinking.
Many people will rely on it and just stop thinking for themselves.
However the genie cannot be put back in the box.

This guy has some interesting thoughts.

 

[chiefdave]

Read an article recently about companies who were early adopters starting to ditch it as they've run the numbers and its taking more time and money to check everything to prevent fuck ups than it is doing it manually in the first place.

Was places like solicitors and accountants who do a lot of receptive admin work, well their assistants do anyway.

Article also mentioned that by adopting AI they'd completely cut off their pipeline of new recruits. This was the work they did which gave them a good grounding to move on to more senior roles but now they're finding that there's nobody to step up when needed.

Sure it will have its uses but people, including the PM, need to stop viewing it as the silver bullet to cure everything.

We also need to think about how we're going to deal with a situation where large numbers of jobs are removed from the economy because they're replaced by AI.

 

[chiefdave]

Many people will rely on it and just stop thinking for themselves.

Most people on twitter already have. Repies to the most basic things filled with grok requests

 

[Brighton Sky Blue]

Most people on twitter already have. Repies to the most basic things filled with grok requests

When I was growing up, search engines already felt kind of a lazy way to find information. Now there's AI summaries of what the search engine will bring up so you don't even have to go checking through the results.

Saw someone on the bus last year using AI to write them a sick note excusing them from going to an interview. What kind of employer would want that anyway?

 

[rob9872]

There’s no doubt that artificial intelligence has the potential for good: it can streamline repetitive tasks, improve efficiency, assist in complex decision-making (for example in healthcare or logistics) and open up new possibilities that were simply impractical before.

At the same time, the risks are real: displacement of jobs, concentration of power among those who control the AI, ethical issues around bias and fairness, and the broader question of how society adapts to major technological change.

My view is that AI isn’t inherently good or bad — it’s a tool, and the outcome depends a lot on how we deploy it, regulate it and embed it in our economy and culture. The biggest challenge (and opportunity) will be in getting the governance and human-systems side right: ensuring that as we adopt AI we do so in ways that protect people, distribute the benefits, and keep sight of the values we care about.

In short: yes — use it. But use it consciously

 

[wingy]

Read an article recently about companies who were early adopters starting to ditch it as they've run the numbers and its taking more time and money to check everything to prevent fuck ups than it is doing it manually in the first place.

Was places like solicitors and accountants who do a lot of receptive admin work, well their assistants do anyway.

Article also mentioned that by adopting AI they'd completely cut off their pipeline of new recruits. This was the work they did which gave them a good grounding to move on to more senior roles but now they're finding that there's nobody to step up when needed.

Sure it will have its uses but people, including the PM, need to stop viewing it as the silver bullet to cure everything.

We also need to think about how we're going to deal with a situation where large numbers of jobs are removed from the economy because they're replaced by AI.

It's quite simple really, just be concentrated downward to a few extremely wealthy individuals ,is that not what happens every revolutionary discovery why do we think they'll be benign?

 

[Sick Boy]

Also it’s better to refer to it as automated machine learning nothing artificial about it

This is not the case. They're pre-trained deep learning models and use retrieval-augmented generation to supplement the data.

 

[Brighton Sky Blue]

This is not the case. They're pre-trained deep learning models and use retrieval-augmented generation to supplement the data.

I think what Google did with Alpha Zero and chess was pretty remarkable and is more what I think of as 'artificial intelligence'. Taught itself the essence of good play and played itself millions of times to learn what worked best before destroying the brute force chess computers.

On the whole though the environmental impact of stuff like Grok is monstrous because of how much power it churns through. I've read some Scandinavian countries have offered companies geothermal heat sources as alternatives though.

 

[rob9872]

For the avoidance of doubt, my response was AI generated. I asked Chat GPT to review AI: Good or bad? and create a balanced response for me, then simply posted above.

Scary thing is that I don't think anyone has noticed, which has merits of saving time, but could ruin discussion on here too. All the more reason to stick to football I guess where responses are obciously real and more emotional.

 

[chiefdave]

There’s no doubt that artificial intelligence has the potential for good

 

[fernandopartridge]

It's neutral really though you might argue that it is terrible regardless due to its environmental impact. Good when used for something socially and / or genuinely economically useful.

The problem we've got at the moment is most use cases are not.

 

[ajsccfc]

The world is burning so that a robot can rip off existing information and still get it wrong

 

[Brighton Sky Blue]

For the avoidance of doubt, my response was AI generated. I asked Chat GPT to review AI: Good or bad? and create a balanced response for me, then simply posted above.

Scary thing is that I don't think anyone has noticed, which has merits of saving time, but could ruin discussion on here too. All the more reason to stick to football I guess where responses are obciously real and more emotional.

I was expecting at least a few people to put it through AI to be fair but yes, I couldn't really tell the difference.

 

[Sick Boy]

The world is burning so that a robot can rip off existing information and still get it wrong

It hallucinates and gets stuff wrong a lot of the time...anyone feeding it data and trusting its calculations might want to start checking.
IMO, we're at the start of the end of the open web.

 

[Brighton Sky Blue]

It hallucinates and gets stuff wrong a lot of the time...anyone feeding it data and trusting its calculations might want to start checking.
IMO, we're at the start of the end of the open web.

I played ChatGPT at chess. After about 10 moves, it forgot where the pieces were, which ones had been taken, that it was in check etc. A bit further on it even wanted to move pieces to squares that don't exist

 

[Sky Blue Pete]

This is not the case. They're pre-trained deep learning models and use retrieval-augmented generation to supplement the data.

Sorry sb I meant more generally thanks for clarification though

 

[Ian1779]

If AI is going to destroy the world, I at least have the confidence they will do it efficiently and quickly.

 

[Otis]

It was very helpful in getting me out of a pickle to be fair to it.

A band asked me to review a new single of theirs and I really didn't like it

Couldn't bring myself to lie about it, so got AI to fib for me on my behalf.

 

[olderskyblue]

I’ve tried to use it to animate some old photos of my dad (he died when I was 17) but the free stuff only gives you a couple of goes, and nothing seemed to work well in that short a time. 1st one changed his face completely, 2nd didn’t understand the movement I asked for. How do you test these things to see what’s worth paying out for, or not…

 

[Otis]

I’ve tried to use it to animate some old photos of my dad (he died when I was 17) but the free stuff only gives you a couple of goes, and nothing seemed to work well in that short a time. 1st one changed his face completely, 2nd didn’t understand the movement I asked for. How do you test these things to see what’s worth paying out for, or not…

My laptop has Copilot and it lets me have about 10 stabs at everything, every day.

It's actually been very accurate for me

Gemini, on my phone, was pretty rubbish.

 

[chiefdave]

I’ve tried to use it to animate some old photos of my dad (he died when I was 17) but the free stuff only gives you a couple of goes, and nothing seemed to work well in that short a time. 1st one changed his face completely, 2nd didn’t understand the movement I asked for. How do you test these things to see what’s worth paying out for, or not…

Was reading article about tests that have been done with missing Doctor Who episodes. They have ‘telesnaps’ which are essentially photos of what was broadcast and also the soundtrack so in theory it would be possible to rebuild the episodes.

Seems the tests show we’re still a long way off that but with the speed things are moving you could only be talking years before something like that is possible.