this post was submitted on 05 Jul 2024
1097 points (95.4% liked)
Memes
45633 readers
1163 users here now
Rules:
- Be civil and nice.
- Try not to excessively repost, as a rule of thumb, wait at least 2 months to do it if you have to.
founded 5 years ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
You dont think polluting the world is going to have a net negative effect for humanity?
What exactly is there to gain with AI anyways? What's the great benefit to us as a species? So far its just been used to trivialize multiple artistic disciplines, basic service industries, and programming.
Things have a cost, many people are doing the cost-benefit analysis and seeing there is none for them. Seems most of the incentive to develop this software is if you would like to stop paying people who do the jobs listed above.
What do we get out of burning the planet to the ground? And even if you find an AI thats barely burning it, what's the point in the first place?
The whole point is that much like industrial automation it reduces the number of hours people need to work. If this leads to people starving then that's a problem with the economic system, not with AI technology. You're blaming the wrong field here. In fact everyone here blaming AI/ML and not the capitalists is being a Luddite.
It's also entirely possible it will start replacing managers and capitalists as well. It's been theorized by some anti-capitalists and economic reformists that ML/AI and computer algorithms could one day replace current economic systems and institutions.
This sadly is probably true of large companies producing big, inefficient ML models as they can afford the server capacity to do so. It's not true of people tweaking smaller ML models at home, or professors in universities using them for data analysis or to aid their teaching. Much like some programmers are getting fired because of ML, others are using it to increase their productivity or to help them learn more about programming. I've seen scientists who otherwise would struggle with data analysis related programming use ChatGPT to help them write code to analyse data.
As the other guy said there are lots of other things using way more energy and fossil fuels than ML. Machine learning is used in sciences to analyse things like the impacts of climate change. It's useful enough in data science alone to outweigh the negative impacts. You would know about this if you ever took a modern data science module. Furthermore being that data centres primarily use electricity it's relatively easy to move them to green sources of energy compared to say farming, or transport. In fact some data centres already use green energy primarily. Data centres will always exist regardless of AI and ML anyway, it's just a matter of scale.