What jobs & university courses should kids aim for to protect employment from Artifical Ingelligence?

[Suncreamweather]

Following on from the Artificial Intelligence thread.. What careers & university courses will still still be necessary for humans to carry out?

These types of jobs require a certain type of ability and skillset. Most ordinary people won't be able to achieve the level of education or have the necessary skills or abilities to perform them.

I think AI could become good at teaching. It would be able to observe learning styles and personalise approaches, something which is more difficult for a human teacher of 35 children at a time. Schools would probably still need human staff to oversee the pupil's welfare. I think it will struggle to replace creative industries such as music, art or literature which often arise out of painful human experiences and the urge to transcend them.

Outdoor education.

There are many jobs and careers that might be present in years to come that haven't been invented yet.

At the time of the industrial revolution, people probably predicted mass unemployment and half the labour force not being needed but new roles appeared.

When factories became more automated...ditto.

When computers meant a lot of clerical staff were no longer needed in such numbers (typing pools, secretaries...)...ditto.

Things evolve.

@Trumporange That's true, but we have a lot more people who are living and working for longer now, and more women in the workforce.

I think AI has the potential to do a bigger range of jobs than previous technology ever has because, as the name suggests, it's intelligent. A steam train, mechanised factory, combine harvester, computer etc. weren't intelligent and still needed people to work them. This time it's going to be different, apparently.

My DS is studying robotics. He reckons there will be work in that field for a while at least!

Can I ask, taking for instance a waiters job. I can see how this could be automated. Easily. We have a robot scoot over to the table to take order, customer tells robot. Or types in to screen. And robot goes away and inserts code to kitchen. Robots in kitchen then make food and robot knows where to collect plates from collection point and deliver to table. I assume this is predicted AI future? But surely it’s just very very expensive for the average restaurant to buy and store all these robots? Surely it will be cost prohibitive at least for a while?

It would probably come in first to big chain restaurants, particularly fast food - McDonalds being an obvious one.

It really depends on the maintenance costs of the robots vs wages paid to the average worker. If they are cheap to maintain they would pay for the initial outlay in saved wages quite quickly - and of course they could work 24/7 365 days a year.

I think you'd still need a human around for customer service issues. Although the robots wouldn't make mistakes in the way tired human workers might, in my experience at least 50% of customer service issues are due to the customer either being wrong (e.g ordered the wrong thing and complained when it arrived) or not liking something ('this new coffee you've brought in is horrible') and they will want to complain to someone who can show genuine, not artificial, empathy.

The automated bank phone calls are certainly pretty awful just now. I tried to phone up an ask a question and didn’t get far at all

Anyone who has seen Humans knows robots need downtime to recharge. The Jueying X20 can walk (max pace 5km/hr) for max 4 hours without a charge. I can't find any info about charge-time, I bet it's at least a few hours. Have you seen price of electricity?

If mischievous kids keep making messes in the diner, what's to stop kids tying up restaurant robot with always cleaning messes & never serving customers?

Thats are really good point about electricity cost. We are going to have to get a lot better at renewable energy if we are going to be able to power all these robots

Yeah, I don't think a robot could replace a teacher in a classroom setting for this reason. Badly behaved kids aren't going to listen to a robot telling them off. It would be chaotic. Although I guess you could get a robot to do the teaching and then just pay a human to be present to oversee behaviour.

Because if you do that in a diner you’re identified by CCTV and your parents don’t get their UBI that month.

The problem with all these “ aha but they won’t be able to build a fence/ be a teacher/ work in a diner” scenarios is they assume AI functioning in a world created by humans with AI as the slavebot. That’s not AI. AI would be able to reconfigure the world. Why would there be a fence when it could think of a better way to solve the problems that a fence currently solves. Don’t ask me what that solution is because how can I with an IQ of 80-140 compete with something with an IQ of 1000+. It would be like me asking a squirrel to design me a car.

Again, you’re basing your opinions on current and past robots. They are nothing like AI and cannot be compared.

You are unable to comprehend AI because we’ve never seen anything like it before. But everything you seem to think they can’t do, they absolutely will be able to.

Is it too late to pull the plug on AI? It sounds as though it will create far worse problems than it solves.

I agree with Trump on Industrial Revolution predictions but in relation to more people now it seems we also are meant to be concerned about population decrease so maybe we’ll get some decrease at the right time

It is hard to get your head around. Up until now computers have largely just collated and replicated human knowledge and done it faster and more accurately because they can run iterations faster and don’t get distracted and get their times tables wrong. We know what they’re doing though.

i recommend reading Scary Smart by the guy who used to head up Google Labs. Despite the title it’s a pretty balanced view of where AI is going and what it is/ isn’t.

What I found really interesting is that AI uses almost a machine version of “eugenics”
in that the developers create a load of bots ( systems not actual robots), set them a task and then weed out the ones that fail and further develop the ones that show they can best “learn”. For example they wanted them to create a trading system and the best ones created a barter system and a value system and actually developed a sense of themselves as separate from and in competition with the other trading bots. It’s the difference between giving the computers instructions for a trading game and asking it to execute ( existing tech) and giving one a pile of apples and one a pile of pears and some videos of people trading things and seeing what they come up with. Now this is obviously the bottom rung of AI but you can see that it’s already a fundamental departure from the if, then model.

Automation is already there in fast food restaurants. Many of them you make your own menu selection and pay at a console rather than talking to a human being. Same in retail self-checkout. We’ve normalised this now and find it unremarkable, but these were jobs done by people only a few years ago.

Jobs are only done by people while they are cheaper than automation. The moment a robot can do the same job more cheaply (they don’t even have to do the job as well, or as quickly) that person is gone. I’m

Construction robots are already a growth market and being used in ‘real life’ (not just as concept robots).

In this video we see robots bricklaying, building drywalls etc. Some completely autonomously some still needing human participation (for now).

..and "robot" delivered room service has been available in some hotels for quite a few years now, but there's much real AI involved - though it is impressive to see them (if you haven't seen them think meter high Roomba's) negotiate things like lifts.

Primary teaching, police, emergency services. Counselling

This is an interesting discussion about AI and AGI https://m.youtube.com/watch?v=GIUhyEK9LJQ&pp=ygUMbm92YXJhIG1lZGlh

I am a bit slow to be noticing all this AI stuff and need to read up on it more. It is scary. When experts are saying that AI will start doing things that we actually don’t understand how it is doing them. At the moment we are all assuming AI is like a slave to humans where their mission is to create a better world for humans. But it seems a point will come where we aren’t actually in control any more? And AI could easily then create a new mission all of their own that we cannot stop?

@Bartg it is very scary.. It really is the great unknown isn't it..