AI Is Coming for Your Job, Much Faster Than Anyone Thought | Hacker News Reader

AI Is Coming for Your Job, Much Faster Than Anyone Thought

decrypt.co

11 pointsmethuselah_in17 comments

For the rest of my life, AI will be coming for my job and full-self-driving cars will be just around the corner.

Don't forget about nuclear fusion! It's the future!

1960s - Herbert Simmons predicts "Machines will be capable, within 20 years, of doing any work a man can do."

1993 - Vernor Vinge predicts super-intelligent AIs 'within 30 years'.

2011 - Ray Kurzweil predicts the singularity (enabled by super-intelligent AIs) will occur by 2045, 34 years after the prediction was made.

The predicted advent of general purpose AI recedes into the future by more than 1 year per year.

The entire point of AI is not to make people more productive, but to trade productivity in the short-term for even more automation that goes beyond the point of diminishing returns so that those at the top can accumulate even more wealth. Big-tech corporations act as attractants for those who love power and money above any other concern and they certainly are not creating any tools to help anyone.

When a small company makes a product, they are happy to make something useful in exchange for a living. When big tech makes something, you are the product and they just want to use their economies of scale to squeeze you dry, making you think they have something useful.

People talk about AI now as if it is a great tool for them. Maybe it is, but in some years it won't really be a great tool at all except at keeping power in the hands of the most wealthy.

Isn't this the case with any invention that greatly improves productivity? Ask early 1800s English textile workers. Their wages did not return back to late 1700s peak, inflation adjusted, till WWI. Does that mean that mechanisation of textile industry was a bad thing? It meant people no longer had to wear $5000 T-shirts (https://www.sleuthsayers.org/2013/06/the-3500-shirt-history-... - I inflation-adjusted $3500 from 2013 to today).

The people alive in 2225 will be happy with what happened, and suitably grateful I am sure.

If it was made in the US with a $7.25/hour wage sure, now do it in cambodia for $200 a month, at 48 hours a week that's barely a dollar an hour, all of a sudden it's a $500 shirt already.

Anyways I don't think you can use inflation over 700 years to come up with a tshirt price based on modern min wages, even peasants had clothes and they sure did not spend 6 months of labor to get a single item of clothing

After spending some times looking about more info on this topic you very quickly find people who debunk the original claim, and the inflation adjusted price the list is about I came up on my side using cambodia: https://www.bookandsword.com/2017/12/09/how-much-did-a-shirt...

It is the case, but it varies in magnitude. The textile industry was a bad thing in one way in that it caused an uncontrollable surplus of clothing that is in a sense "too cheap" and now is a major environmental problem. They key is that each invention can be thought of having a magnitude M of how it becomes uncontrollable, and once M > T (some threshhold T), then it becomes a net negative, where AI is certainly greater than T.

I probably agree with that one. Indeed, it's a bit of a "too much of a good thing" by now.

I don't know whether it's coming for my job or not, but I've found it incredibly useful for dealing with situations I'm completely unfamiliar with.

For example, I have a small hobby business with a few friends. One of them suggested that we put up a stall at a major upcoming exhibition.

I've never done one of these before, so I asked GPT o3 to help me figure out what instructions to share with my fabricator. The output was so good that the fabricator said I'd made his job 5x easier.

Once I got the 3d render, I asked o3 to help me figure out how to draw more attention. It gave me display ideas for different budgets, copy ideas, and layout ideas.

The display ideas were fantastic and it would have taken a few hours of brainstorming to come up with them.

The copy ideas were also really good. I've worked as a copywriter in the past and gun to my head, I could maybe get about 10% more juice out of them. For the average business, it was perfectly fine at first pass.

If this is true, can someone even remotely explain why boards and companies don’t see the existential risk this would pose to the economy? When the people that buy things and other companies that buy things can’t buy anymore, who gains from it?

Same reason a lot of governments seem to do completely counter productive things: short term personal gains, by the time the consequences are visible they're already retired with a fat stack of cash in the bank

The richest already have nearly 50% of buying power. Elon Musk can make cities building rockets the size of skyscrapers. You don’t have to look forward, you can look at the past robber barons for examples. Yes it’s bad for the economy but it’s great for a few people who acquire wealth.

The government is reactive, not proactive

> the existential risk this would pose to the economy

Yes, the existential risk this would pose to the economy is that you sudenly need 100 people to clean up the mess left by AI.

When OpenAI has an outage - all your colleagues just stop working at once. The price will go up, and you will be stuck, when half your teams are machines. Meanwhile your customers are out of work, and your business tanks. The problem is not that this will happen, the problem is your boss thinks this will happen and is no longer hiring.

Dear Author, AI is coming for YOUR job. The doomsday marketing approach is getting rather old, IMO.

Reply on news.ycombinator.com