As Companies Embrace AI, It's a Job-Seeker's Market (reuters.com) 65
An anonymous reader shares a report: Artificial intelligence is now being used in an ever-expanding array of products: cars that drive themselves; robots that identify and eradicate weeds; computers able to distinguish dangerous skin cancers from benign moles; and smart locks, thermostats, speakers and digital assistants that are bringing the technology into homes. At Georgia Tech, students interact with digital teaching assistants made possible by AI for an online course in machine learning.
The expanding applications for AI have also created a shortage of qualified workers in the field. Although schools across the country are adding classes, increasing enrollment and developing new programs to accommodate student demand, there are too few potential employees with training or experience in AI. That has big consequences. Too few AI-trained job-seekers has slowed hiring and impeded growth at some companies, recruiters and would-be employers told Reuters. It may also be delaying broader adoption of a technology that some economists say could spur U.S. economic growth by boosting productivity, currently growing at only about half its pre-crisis pace.
[...] U.S. government data does not track job openings or hires in artificial intelligence specifically, but online job postings tracked by jobsites including Indeed, Ziprecruiter and Glassdoor show job openings for AI-related positions are surging. AI job postings as a percentage of overall job postings at Indeed nearly doubled in the past two years, according to data provided by the company. Searches on Indeed for AI jobs, meanwhile increased just 15 percent.
The expanding applications for AI have also created a shortage of qualified workers in the field. Although schools across the country are adding classes, increasing enrollment and developing new programs to accommodate student demand, there are too few potential employees with training or experience in AI. That has big consequences. Too few AI-trained job-seekers has slowed hiring and impeded growth at some companies, recruiters and would-be employers told Reuters. It may also be delaying broader adoption of a technology that some economists say could spur U.S. economic growth by boosting productivity, currently growing at only about half its pre-crisis pace.
[...] U.S. government data does not track job openings or hires in artificial intelligence specifically, but online job postings tracked by jobsites including Indeed, Ziprecruiter and Glassdoor show job openings for AI-related positions are surging. AI job postings as a percentage of overall job postings at Indeed nearly doubled in the past two years, according to data provided by the company. Searches on Indeed for AI jobs, meanwhile increased just 15 percent.
Re: (Score:2)
Bring it on! I'll pit their garage-sale used pitchforks against my Boston Dynamics dancing kill-bot any day.
Re: Boosting productivity (Score:1)
Interesting field, but.. (Score:1)
The math behind it, to me, is impenetrable. I think for many people, without the necessary math background, learning AI will be difficult unless you're comfortable working with a black box, although I suspect that would only take you so far.
Re:Interesting field, but.. (Score:4, Interesting)
Re:Interesting field, but.. (Score:5, Funny)
I will just wait for the 1 week bootcamp AI certification class...
Re: (Score:1)
Re: (Score:1)
Not-having-to-work (i.e. losing jobs) can be viewed as our goal within all economic systems. No matter where you are on the spectrum of Adam Smith to Karl Marx, our time above-ground is a scarce resource. Every-fucking-thing that is expensive, is ultimately expensive because it used up someone's time, where that person sighed and walked a few more steps toward their dusty, eternal grave, working on your whatever, instead of living their life. The dollars are just a measurement of how much life you as
Re: (Score:1)
Obviously the trucker must become an AI developer. New jobs for everyone! ... This just in! Higher IQ people do not understand not understanding by lower IQ people. More at 11 as the situation develops!
Counter-point (Score:2)
Re: (Score:1, Insightful)
I'm 36 - it's been a lot of work but ramped up at a company using machine learning in a very specific space - but gives me the time to ramp up and educate myself on techniques and everything else.
Stay current or die, it's always been true.
Good luck out there.
Re: (Score:3, Informative)
Re:Counter-point (Score:5, Informative)
You are somewhat correct, if esoteric research in the 1960's is what you are referring to as 'known techniques'. Modern AI techniques weren't truly implemented until the mid 70's, with broader acceptance and applications demonstrated in the mid 80's: https://en.wikipedia.org/wiki/... [wikipedia.org]
If you are referring to the 1960's symbolic (lisp) techniques espoused by the disgraced Marvin Minsky [wikipedia.org], nearly fired from MIT for borderline fraud, a case that saw MIT forced to repay DARPA millions in wasted research money, then you are a bit behind the times. Modern AI techniques are now quite far from Minsky's self-aggrandized approach. Modern techniques were pioneered more by Minksy's high-school rival and a victim of Minky's petulant personal and private bullying, the truly brilliant Frank Rosenblatt [wikipedia.org]. Frank was so close. Had he lived just a few more years and kept his confidence, he would have seen his dream realized.
Why it took the span of a human lifetime for people to see through Marvin is baffling: https://www.reddit.com/r/Machi... [reddit.com]
Anyway, modern AI takes a bit of calculus to truly understand, and some statistics. An undergraduate with a solid math foundation should be able to derive the backprop algorithm and explain it. Then there's catastrophic forgetfulness, SLAM techniques with grid and place cells... probably things beyond a typical undergraduate curriculum, but possible.
Any technician can be trained to push buttons. It might take a bit more fundamental understanding of what is going on under the covers to catch training pitfalls and prevent inefficiencies. Maybe this is what companies hiring for AI work are after.
Re: (Score:2)
Here I go again, replying to an AC...
Keep digging, you'll find it. You might have to actually look at archived printed media, because MIT and other really don't want that part of their history known, and they spend a bit of time and money polishing their image.
What forced Minsky into the Icon Media Lab? (He didn't go by choice) Who was his fellow MIT professor that was fired?
Frank Rosenblatt's official cause of death is listed as a boating accident, but was it?
When mob rule decides what makes it into the
Re: (Score:2)
The technique could best be called interconnecting branched prediction techniques (new branches can be formed and connect to existing branches), the predictions could be considered leaves, taking in and putting out data. Each branch a relational data construct, processing the data incoming and outgoing and of course the trunk is the collation point for all data and the branch constructor. An AI, well but then we can be programmed too, so, hmm, the more complex the AI the more unreliable and unpredictable it
Re: (Score:2)
PS the roots, the hardware.
Re: (Score:2)
Re:Counter-point (Score:4, Insightful)
Automation creates more jobs than it displaces in the long run. Jobs change, and you have to be prepared to adapt, learn new things, and possibly train for a new career, otherwise, you're just trying to perpetuate a conservative system.
He's 41. Let's say it takes 5 years for these :new careers" to open up, then another 5 years for him to train up. He's now 51, with no experience in that new career, but all the attendant demands/requirements that tend to come with 51 year olds-kids in or about to enter college, desire for work/life balance, reasonable pay and benefits especially healthcare. Who is a company more likely to hire? Him, who will also probably like to retire in 10-15 years, or a 20-something fresh out of college with the exact same amount of experience in that career field who is cheaper both in absolute terms (lower salary) and more intangible terms (less likely to need/use healthcare, fewer sick days, fewer to none family obligations, etc).
Will a lucky few get hired for a reasonable wage? Sure. Will a few more be willing to take much lower wages and find a company willing to hire them? Probably. Will a lot more get kicked to the curb and told "tough luck, try applying at Walmart, Home Depot, or an Amazon warehouse"? Most definitely.
Re: (Score:2)
"retire in 10-15 years"
i will just mention that the average time span per company in US is ~4.5 years with older staying longer.
Re: (Score:2)
Re:Counter-point (Score:5, Interesting)
AI requires a lot of education and quite a background in technical expertise.
Actually ... it doesn't. Deep learning uses a lot of linear algebra, differential equations, and complicated algorithms to deal with regularization and efficiency. But all that is tucked away in libraries. For a real-world AI app, you just slap together a Tensorflow pipeline using Python, and fiddle with the parameters until you get good results. It is more art than science.
This is how it works [xkcd.com].
My son is 15, and he went to an "AI bootcamp" this past summer. It was a two week course, and he built a pretty snazzy reinforcement learning application, using Python and some canned visualization tools. Later he made a generative NN to create animations. This is a kid that is just starting high school.
It's certainly out of reach for me at 41 years old.
Probably, but because of your attitude, not your age.
I really hate the whole concept of AI because it is putting people out of work
There is zero evidence that AI is "putting people out of work". How many people do you know that have lost their jobs to deep learning?
What this sounds like is Late Stage Capitalism.
You should spend more time on professional development, and less time reading The Communist Manifesto.
Re: (Score:2)
AI requires a lot of education and quite a background in technical expertise.
Actually ... it doesn't. Deep learning uses a lot of linear algebra, differential equations, and complicated algorithms to deal with regularization and efficiency. But all that is tucked away in libraries. For a real-world AI app, you just slap together a Tensorflow pipeline using Python, and fiddle with the parameters until you get good results. It is more art than science.
This is how it works [xkcd.com].
If I'm reading the XKCD correctly, there will be a lot of "AI" jobs for those of us that actually know the underlying math, even if it means we're closer to 41 than 15.
Re:Counter-point (Score:4, Insightful)
If I'm reading the XKCD correctly, there will be a lot of "AI" jobs for those of us that actually know the underlying math, even if it means we're closer to 41 than 15.
Indeed. Like any other field, AI will bifurcate into "tool builders" and "tool users", with the former being much better paid, and the latter being far more numerous.
To code an AI engine you may need to know how to transpose a Jacobian matrix. To use it, you just import a library.
Disclaimer: I am a lot older than 41.
Re: (Score:1)
For those of us who do understand AI at a deeper level and who do have a sound mathematical background, and have built neural nets, genetic algorithms, and many others from the ground up to drastically increase competitiveness of companies in the real world it's not entirely clear where the jobs are still.
Because, at the end of the day, I can still get paid more building dull ASP.NET MVC applications than I ever could doing any of the stuff above, so as much as I'd love to do AI again professionally, frankl
Re: (Score:2)
Re: (Score:2)
If I'm reading the XKCD correctly, there will be a lot of "AI" jobs for those of us that actually know the underlying math, even if it means we're closer to 41 than 15.
Yep, but you better be able to deal with the bright young punks, or they'll put you out to pasture :)
Re: (Score:2)
I personally specialize at stirring the pile, its been a good career and will likely continue. Much interesting question is what happens to the population as the necessity of cheap labor is lost.
Re: (Score:2)
Much interesting question is what happens to the population as the necessity of cheap labor is lost.
The steel plow and McCormick reaper eliminated nearly all farming jobs at a time when 80% of the population worked in agriculture. We survived.
Deep learning automates image recognition and natural language processing at a time when 0.001% of the population work as image recognizers and transcribers. I think we will muddle through.
Most human jobs require general intelligence. This is completely out of reach for current AI, and there is no clear path in that direction. Machines will get much better at ima
Re: (Score:2)
It's certainly out of reach for me at 41 years old.
Probably, but because of your attitude, not your age.
Perfect. Thank you for that.
Your son will not push the state of the art, and will likely not be able to keep up in competitive new markets, with just the 'use a library' mentality. But you already know that, as your other comments hint at it. A bit of formal education in AI will help anyone, and is not out of reach for anyone that wants to learn.
Your son sounds like a very bright kid who probably has a very bright mentor to encourage him along. Well done!
Re: (Score:2)
Your son will not push the state of the art, and will likely not be able to keep up in competitive new markets
Why would anyone want to do that? Why would anyone want to play the startup-hopping coke-fueled game of working 90 hours a week to be "cutting edge competitive" and be ousted at 30 years old?
Only fools travel that road. "Settle" for an easy 40 hour week at 120k and a 3k sq. ft. house in Austin. Get a banging wife to bang and enjoy your life. Go on vacation several times a year. Have some fun. Know joy outside a leased office space. Get to know your kids better than your boss. Let the idiots have the "joy" of pushing the state of the art.
I could live with that.....
Re: (Score:2)
McDonald's is also cashierless in parts of Europe and the U.S. due to touchscreens.
Touchscreen menus are not "AI".
Again, no one lost jobs, but jobs just weren't being created.
Every dollar not spent on employing a cashier is a dollar spent (or invested) elsewhere in the economy, generating other jobs. But these new jobs are likely to involve actual production of goods and services, unlike the pointless and unproductive cashier. Since the point of jobs is production of goods and services, and not "keeping people busy", this is a benefit to the economy and makes us all better off.
Economies are not "zero sum": Lump of Labor Fallacy [wikipedia.org]
Re: (Score:1)
You can transition at any age (Score:2)
AI requires a lot of education and quite a background in technical expertise. It's great that there are many openings but the path to obtain the job is out of reach for a lot of people. It's certainly out of reach for me at 41 years old.
This is pretty distressing to read, as it's really not at all the case. Do not think you could not enter this field - I started to transition to machine learning work last year, and am older than you are.
I would recommend taking some actual course to see if the work even in
Unfortunately (Score:2)
They hire only Artificial Intelligence, no wetware wanted.
High bar for entry (Score:1)
Awww...this is simple... (Score:2)
Ivor (Score:2)
As one of my favorite Computer Scientists, Ivor Paige, once put it, "there's more A than I in AI". What we're today calling AI is still limited-domain expert systems. True AI is still a ways off.