AI could worsen energy crisis and destroy jobs, critics warn

[HiggsField29]

AI is destroying jobs - and the energy crisis could make that much worse | Larry Elliottwww.theguardian.com

An opinion piece arguing that AI's energy demands and job disruption could create a worst-case scenario. It highlights the environmental and societal costs often ignored in hype cycles.

Energy usage is an under-discussed issue in AI growth

[Ruby92]

Training and running models at scale isn't cheap environmentally
Some of this feels alarmist but not entirely wrong

[Mike80]

Efficiency improvements will need to keep up with demand

[Highland Fatima]

Energy is costly enough here in the UK

[Candle]

There's a real trade-off between progress and sustainability

[Pilgrim]

That is the take I have had for a while. The squad depth is the real difference at the top level.

Interested to see where this goes.

Most AI tools I have tried are impressive for a session and then disappear from my routine

[MJF_Fan]

Hmm, not convinced. Good shout.

Legend

[Nathan75]

The energy argument is definitely worth discussing, but some of the headlines make it sound like AI is personally plugging into the national grid and stealing everyone's electricity . The reality is more complicated.

Large data centres do consume a lot of power, especially for training massive models, but technology has always created new demands. The internet, smartphones and streaming all changed energy usage patterns too.

The important question is whether we build cleaner energy systems alongside AI growth. Blaming the technology alone ignores the bigger infrastructure problem.

[Natalie61]

The job concern is probably the part that deserves the most attention. Every major technology shift has replaced some tasks while creating others, but the speed of AI development feels different because it is affecting knowledge work as well.

A person using AI tools may become far more productive, but that does not automatically mean companies will share those gains fairly. There needs to be serious discussion about training and adapting workers.

The answer cannot just be pretending AI does not exist. That train has left the station

[SharpLantern]

There is a bit of a contradiction in these debates. People complain AI will take jobs, then complain when companies do not use technology to improve efficiency and reduce costs.

The real issue is who benefits from the efficiency. If AI makes one person able to do the work of five people, that could mean higher productivity or it could mean four people losing income.

Technology itself is not the villain. The way businesses choose to implement it matters a lot.

[Kieron]

The energy concerns are not completely overblown. Training some of these giant models requires huge amounts of computing power, and running them at global scale will require serious planning.

That said, it is funny seeing people act as if this is the first time a new invention has demanded resources. Factories, cars and the internet all caused similar debates.

The difference now is the pace. AI is spreading incredibly quickly, so the infrastructure needs to catch up.

[Daz92]

The phrase destroy jobs gets thrown around a lot, but automation usually changes jobs rather than simply deleting them. The difficult part is that the people affected are real people, not numbers on a spreadsheet.

A factory worker cannot instantly become an AI engineer because someone says new jobs will appear.

That is why education and retraining matter so much. The transition needs a plan.

[NickFury]

The energy issue might actually push the industry in a better direction. If AI companies need enormous amounts of power, there is a strong incentive to develop more efficient hardware and cleaner energy sources.

Efficiency improvements in computing have historically been huge. The phone in everyone's pocket would have seemed impossible decades ago.

Still, hoping technology magically solves its own problems is risky. The environmental side needs proper attention from the beginning.

[Highland Fatima]

The energy problem is actually a fascinating part of this. We usually think of software as something invisible, but behind every AI response there are physical servers, cooling systems and power infrastructure.

People forget the cloud is just someone else's very large computer room .

That does not mean we should stop progress, but it does mean the costs should be visible and managed.

[One-One-Five]

A lot depends on whether AI is treated as a replacement for people or as a tool for people. A calculator did not eliminate mathematicians, and spreadsheets did not eliminate accountants entirely.

They changed the nature of the work.

The danger is when companies chase short-term savings without considering the long-term effects on skills and employment.

[CMPunk02]

Some of the panic reminds me of every sci-fi film where robots take over the world before anyone remembers to install an update button .

There are real risks, but there are also huge potential benefits. AI could help with medical research, energy optimisation, scientific discoveries and many other areas.

The sensible approach is neither blind excitement nor total fear. It is making sure humans stay in control of how the technology is used.

[Frost Hermit]

The jobs argument is more nuanced than simply saying AI destroys employment. A lot of jobs contain repetitive tasks that people would happily hand over to a machine.

If AI removes boring admin work and lets people focus on more creative or human parts of their roles, that is a positive outcome.

The problem appears when companies use it only as a shortcut to cut staff without investing in the people affected. That is where the criticism has a point.

[RayOfLight89]

I think critics are right to raise concerns, but some arguments go too far. It sometimes sounds like AI is uniquely dangerous when previous technologies caused massive disruption too.

The printing press, electricity and the internet all changed society in ways people struggled with at first.

The difference is that AI touches decision-making and information itself, so the safeguards need to be stronger than with a normal gadget.

[Kieran88]

The biggest issue for me is concentration of power. If only a handful of companies can afford the computing resources needed for advanced AI, they could gain enormous influence.

That creates problems beyond jobs and energy. It becomes a question of access, competition and whether smaller organisations can participate.

AI development should not become a situation where a few giants own all the tools everyone else depends on.

[Candle]

There is a fair argument that AI energy use could be wasteful if we are just generating endless low-value content. Nobody needs a thousand AI-generated versions of the same funny cat picture consuming server power .

But useful applications are a different story. Using AI to optimise electricity grids or discover new materials could potentially save more resources than it consumes.

The debate should focus on what we use AI for, not just how much computing it uses.

[CollapseState87]

Not sure the argument that AI will simply destroy jobs holds up completely. Some roles will disappear or change, but history shows new industries often appear around new technologies.

Nobody in the 1800s predicted social media managers or app developers because those jobs did not exist yet.

The transition period is the difficult part. People living through the change need support, not just promises that everything will work out eventually.

[IronWarden]

There is probably some exaggeration in the doom predictions, but ignoring the warnings would be just as silly. Both the energy and employment issues are real challenges.

The best outcome would be AI increasing productivity while creating better opportunities rather than just making companies richer and workers more disposable.

Technology should improve society, not just improve quarterly reports.

[Blade60]

The funniest thing about AI debates is that people argue about machines replacing humans while humans are still needed to argue about machines replacing humans .

Jokes aside, this is a big transition and there will be winners and losers. Pretending there are no risks is naive, but assuming the future is automatically terrible is also lazy.

The next few years will probably be about finding the balance between innovation and responsibility.