Should Workers Start Learning to Work With AI? (msn.com) 60
"My boss thinks AI will solve every problem and is wildly enthusiastic about it," complains a mid-level worker at a Fortune 500 company, who considers the technology "unproven and wildly erratic."
So how should they navigate the next 10 years until retirement, they ask the Washington Post's "Work Advice" columnist. The columnist first notes that "Despite promises that AI will eliminate tedious, 'low-value' tasks from our workload, many consumers and companies seem to be using it primarily as a cheap shortcut to avoid hiring professional actors, writers or artists — whose work, in some cases, was stolen to train the tools usurping them..." Kevin Cantera, a reader from Las Cruces, New Mexico [a writer for an education-tech compay], willingly embraced AI for work. But as it turns out, he was training his replacement... Even without the "AI will take our jobs" specter, there's much to be wary of in the AI hype. Faster isn't always better. Parroting and predicting linguistic patterns isn't the same as creativity and innovation... There are concerns about hallucinations, faulty data models, and intentional misuse for purposes of deception. And that's not even addressing the environmental impact of all the power- and water-hogging data centers needed to support this innovation.
And yet, it seems, resistance may be futile. The AI genie is out of the bottle and granting wishes. And at the rate it's evolving, you won't have 10 years to weigh the merits and get comfortable with it. Even if you move on to another workplace, odds are AI will show up there before long. Speaking as one grumpy old Luddite to another, it might be time to get a little curious about this technology just so you can separate helpfulness from hype.
It might help to think of AI as just another software tool that you have to get familiar with to do your job. Learn what it's good for — and what it's bad at — so you can recommend guidelines for ethical and beneficial use. Learn how to word your wishes to get accurate results. Become the "human in the loop" managing the virtual intern. You can test the bathwater without drinking it. Focus on the little ways AI can accommodate and support you and your colleagues. Maybe it could handle small tasks in your workflow that you wish you could hand off to an assistant. Automated transcriptions and meeting notes could be a life-changer for a colleague with auditory processing issues.
I can't guarantee that dabbling in AI will protect your job. But refusing to engage definitely won't help. And if you decide it's time to change jobs, having some extra AI knowledge and experience under your belt will make you a more attractive candidate, even if you never end up having to use it.
So how should they navigate the next 10 years until retirement, they ask the Washington Post's "Work Advice" columnist. The columnist first notes that "Despite promises that AI will eliminate tedious, 'low-value' tasks from our workload, many consumers and companies seem to be using it primarily as a cheap shortcut to avoid hiring professional actors, writers or artists — whose work, in some cases, was stolen to train the tools usurping them..." Kevin Cantera, a reader from Las Cruces, New Mexico [a writer for an education-tech compay], willingly embraced AI for work. But as it turns out, he was training his replacement... Even without the "AI will take our jobs" specter, there's much to be wary of in the AI hype. Faster isn't always better. Parroting and predicting linguistic patterns isn't the same as creativity and innovation... There are concerns about hallucinations, faulty data models, and intentional misuse for purposes of deception. And that's not even addressing the environmental impact of all the power- and water-hogging data centers needed to support this innovation.
And yet, it seems, resistance may be futile. The AI genie is out of the bottle and granting wishes. And at the rate it's evolving, you won't have 10 years to weigh the merits and get comfortable with it. Even if you move on to another workplace, odds are AI will show up there before long. Speaking as one grumpy old Luddite to another, it might be time to get a little curious about this technology just so you can separate helpfulness from hype.
It might help to think of AI as just another software tool that you have to get familiar with to do your job. Learn what it's good for — and what it's bad at — so you can recommend guidelines for ethical and beneficial use. Learn how to word your wishes to get accurate results. Become the "human in the loop" managing the virtual intern. You can test the bathwater without drinking it. Focus on the little ways AI can accommodate and support you and your colleagues. Maybe it could handle small tasks in your workflow that you wish you could hand off to an assistant. Automated transcriptions and meeting notes could be a life-changer for a colleague with auditory processing issues.
I can't guarantee that dabbling in AI will protect your job. But refusing to engage definitely won't help. And if you decide it's time to change jobs, having some extra AI knowledge and experience under your belt will make you a more attractive candidate, even if you never end up having to use it.
I asked ChatGPT that (Score:2)
It said no. So the answer is yes.
Re: (Score:2)
AI is too unpredictable to be good in a productive environment.
Add to it that it has the potential to reveal corporate secrets as it learns over time and can re-use that knowledge to someone else.
Re: (Score:2)
But did ChatGPT cite Betteridge?
But me thinks the answer is "Ask again later" per the magic eight ball. By which time it will be too late to worry about it.
Words of wisdom (Score:4, Interesting)
The best part to leave with is his response: "And if you decide it's time to change jobs, having some extra AI knowledge and experience under your belt will make you a more attractive candidate, even if you never end up having to use it."
Re: (Score:3)
Worriers need to stop freaking out and just figure out what it can do for them.
And if they find that it's not helpful to them right now, there's no point in learning something about it for the future -- because it's going to change. If it ever achieves its full promise there will be no need to learn how to use it, because it will learn how to work with us.
Well, assuming it doesn't kill us all.
Re: (Score:2)
That's some really good advice as to how most people who don't really know where this is all going should approach AI. Worriers need to stop freaking out and just figure out what it can do for them. One thing is for sure, it's not going away.
Fads always eventually go away only to be replaced with something more annoying than the last shiny object.
"And if you decide it's time to change jobs, having some extra AI knowledge and experience under your belt will make you a more attractive candidate, even if you never end up having to use it."
I'm not so sure this would be universally viewed positively. I can imagine a programmer, lawyer or artist listing their experience with AI and being rejected from consideration because of it.
Re: (Score:2)
Re: (Score:3)
until the AI bubble bursts
The bubble will burst because of a failure to monetize, not a failure of the underlying technology.
People are using AI for free. Why will they start paying hundreds of billions for it?
It was the same in the Dotcom crash. Pets.com had a failed business model, but the Internet didn't go away.
Re: Words of wisdom (Score:2)
Pets.com failed because of some poor management decisions and bad timing - there was nothing at all wrong with the business model. Everyone buys pet supplies online today.
Itâ(TM)s insane how badly people misunderstand the âoedot com crash.â Nothing crashed but a few management teams and some stock prices. Both the technology and business models proved successful and became ubiquitous in a few short years.
People using the Internet to claim AI will fail just like the internet did is a special k
Re: Words of wisdom (Score:2)
Exactly this:Learn what it's good for (Score:2)
Many of the rest of my prompts seem to be "I have this code..."
One of my pet-peeve with humans is getting a different answer when you ask the same question twice and that's a h
You need critical thinking skills (Score:4, Insightful)
This like saying people should learn programming. You can't learn programming if you don't have a firm grip on algebra. Programming is a bunch of symbols that represent other things. That's algebra.
Unleashing idiots with AI... well... we're already here.
Re: (Score:2)
That is completely untrue. Hell, I don't think any software I ever wrote required any understanding of algebra at all, outside of academic assignments.
Re: (Score:2)
Really? Every time you you assign a common sub expression to a variable, for efficiency or readability, you are using algebra, whether you realize it or not.
Re: (Score:1)
Sure, buddy. Sure.
Re: (Score:3)
You just do not know what "algebra" means.
Re: (Score:3)
Most people don't have critical thinking skills or any logic, so the answer remains a firm NO.
Agreed. If you cannot fact-check AI, it will screw you sooner or later.
The 1990s called... (Score:1)
Love it or hate it, AI isn't going anywhere. In between the "AI is going to make the world a Utopia" crowd and the "AI is just a fad" folks there's a whole lot of people using it every day to make their life easier. You're either going to figure out how to leverage it to your advantage, or get left behind.
Re: (Score:3)
I do not use AI. And I don't feel like I'm being "left behind".
FOMO is not a good reason to do something.
Re: (Score:2)
I don't feel like I'm being "left behind".
Neither did the buggy-whip makers. At first.
FOMO is not a good reason to do something.
Agreed. But ignoring a useful tool just because you don't like it, or whatever your reason for not wanting to use it, isn't a great plan either. If you honestly believe AI is going away I'll take that bet any day of the week.
Just like any other tool you have to understand its capabilities and limitations. I don't expect AI to write me a new novel, paint me a masterpiece, make me lunch, or write the next version of Excel. But it's really good at re-wording e-ma
Re: (Score:3)
I've played around with AI. I was unimpressed. I don't really know what it means to "learn" AI. As far as I can tell, you just hector it and keep rewording what you ask it until the turds it spews out are minimally smelly and maximally acceptable.
There's no science to this. Given that an integral part of every LLM is a random number generator, I don't see how there can possibly be any science to it.
Re: (Score:2)
Lots of useful things have a random number generator as an essential component; that's not a meaningful criticism. It's bound by statistics, so that gives you a useful sandbox to work within.
Indeed, a lot of what you're doing is asking it to give you an idea of what's going on and then fact-checking it. Part of 'learning AI' is understanding what it is and isn't good at, and the kinds of prompts that are likely to get you useful answers.
For instance: I use ChatGPT now and then to help me diagnose and fix te
Re: (Score:2)
Part of 'learning AI' is understanding what it is and isn't good at, and the kinds of prompts that are likely to get you useful answers.
OK. But that doesn't directly help you get your answer, and so doesn't directly help your productivity. And in my experience, figuring out the kinds of prompts likely to get you useful answers is trial-and-error and a bit of a black art, so I'm not convinced that AI will ever meaningfully improve people's productivity. Instead of being stuck trying to think of an answ
Re: (Score:2)
To a certain extent, I don't disagree. I think programming via LLM is an insane goal, and frankly will just lead back to something akin to a programming language, just more poorly defined. Learning some arcane incantation to get consistently good answers isn't anything I would stake a career on.
But it really HAS been useful to me when I'm trying to learn new programming languages because the state of documentation is so poor. As long as I have links back to source documentation, I can read up if something g
Re: (Score:2)
I don't really know what it means to "learn" AI.
The same thing it means to learn how to properly operate any other tool. Learn what it does. Learn what it doesn't do. Learn how to get it to do what it's supposed to do. I'm genuinely confused as to why that's not a straightforward concept. Have you not learned techniques to make a more effective google search prompt? It's literally the same idea.
Middle-Earth Management AI Challenge. (Score:5, Funny)
"My boss thinks AI will solve every problem and is wildly enthusiastic about it," complains a mid-level worker at a Fortune 500 company..
Sit your boss down with AI and start asking it those every problems then.
Start with a nice friendly challenge. Something along the lines of "Can micro-managing middle-earth management cube farmers be easily replaced by AI, with significantly more cost savings due to outdated executive compensation models?"
I'd estimate the corporate-wide *.ai firewall block will be in place..by the end of this sentence.
Re: (Score:3)
Re: (Score:3)
Can micro-managing middle-earth management cube farmers
You have cubicles? Christ I hate people bragging on the internet.
Re: (Score:2)
Boss's job is next (Score:2)
It's not long until the AI is the CEO of companies - either literally or just with a human frontman to sign legal documents.
Re: (Score:2)
That is never going to happen. Or if it happens, it will be about as short as it can get.
Using a tool outside of what it can do is just one thing: dumb.
Managers should get AI training (Score:2)
Or be replaced with AI altogether.
Same as it ever was (Score:4, Insightful)
Management gets excited at the prospect of exploiting more labor. They especially love it if a little curtain can be drawn around the unethical practices of training AI on unlicensed IP.
My timing was excellent (Score:5, Insightful)
I retired in April, 2023. Right before all this AI BS exploded and turned software development into a hellhole. As long as AI doesn't destroy humanity or tank the stock markets, I consider myself extremely lucky to have escaped the AI hype.
The AI snake-oil peddlers are pushing hard. But there's yet to be a single company making profit from AI, save Nvidia which profits by selling hardware to the suckers. Not to mention that the gen-AI industry is theft of intellectual property on an industrial scale; a rapacious environment-destroying energy consumption beast; a supremely confident liar; and an exploiter of underpaid workers whose jobs are to make sure the training doesn't go off the rails and filter out the most disgusting or misleading data from the giant training set.
Re: (Score:2)
Good for you.
As to profits, yep, nobody is making a profit with LLMs and the investments are crazy. For most, this cannot even end well anymore if LLMs suddenly become competent. As LLMs remain the dumb statistical parrots they have been all along, the only ones getting out of the hype relatively undamaged are those with small investments. All others are screwed.
Re: (Score:2)
Re: (Score:2)
" As long as AI doesn't destroy humanity or tank the stock markets, I consider myself extremely lucky to have escaped the AI hype. "
I retired in 2021. AI tanking the stock market, and destroying the United States' ability to issue bonds, and canceling social security payments and medicare is actually the thing I fear the most.
Something bad is going to happen. You can see foreshadowing of it all over the place. Can you do something about it as an average person? No. No matter how you are invested, you are pr
Is the onus on your or your boss? (Score:2)
Every technology that bosses think is useful, they pursue and *implement* by integrating it into their business process. So Ford implemented assembly lines for the Model T. Companies implemented virtual machines in data centers. Google implemented Kubernetes-like software across its internal operations (?)
Why is implementing AI technology your headache? Especially if the only interface you have is a prompt? Implementation is your boss's job.
I think certain development tools have integrated AI well. The busi
No (again) (Score:2)
Re: (Score:2)
Tools have always augmented work (Score:2)
Not replaced it.
The LLM hype has two use cases. Search and Summarize, Generate content
Producing content has never been a problem, internet is full of it.
Search and summary provide better access to knowledge. My instinct is that easy to produce, superficial, unabsorbed knowledge is not enough for innovation, but it can provide sign posts and abstracts, that's useful like a teacher is. One thing LLM cant do is original content, it will always be derivative. I doubt Einstein, Da Vinci, Bach are considered deri
Re: (Score:2)
> Search and Summarize, Generate content
> Producing content has never been a problem, internet is full of it.
LLMs are not good at de novo content generation (vs summarization etc) in general, but one exception is coding - they are good at generating code for boilerplate and simple repetitive tasks, vibe-coded prototype or throwaway code where quality doesn't matter, and can also be used for more serious software development if used appropriately - not "write me a program to do X" vibe coding, but as a
Re: (Score:2)
AGI feels like a bit fusion, except we at least we know what fusion looks like.
The Turing test is bust by LLM. How do we reliably measure general intelligence? What we see in the natural world? Seems a crucial part of that is being able to manifest physically. There are some robots surprising us with how they learn to move, but they are not HAL or ORAC by a long stretch, still operating in very narrow problem domains.
We also don't understand the relationship of consciousness to what we experience as Intell
We will adapt to the result (Score:2)
We used to have newspapers, magazines and websites that looked nice.
Then desktop publishing came along (uuuuh) , a new type of software, and now we have pages full of Hurenkinder (Widow), Schusterjungen (Orphan), Fliegenschiss (Runt), Rivers (in justified text), Bad Hyphenation (consecutive hyphens), Inconsistent Kerning/Tracking, Over-Justification, Inconsistent Leading, Inconsistent Margins, Incorrect Pagination (odd page on left), Misplaced Running Heads/Folios, Lack of Visual Hierarchy, Cluttered Pages
Re: (Score:2)
I was talking about the 'progress' of journalists doing desktop publishing resulting in shitty looking newspapers and millions of job losses.
I really don't get where you got the screen thing from.
I could strangle everybody who forces me to 'next page' an online article article, don't start me on the rest.:-)
Of course (Score:2)
> Should Workers Start Learning to Work With AI?
Well, duh, yes ... AI is here to stay, and will only get better and more useful, even if it remains a very "jagged" capability - great for somethings, and entirely useless for others.
Part of the skill is therefore to understand and learn what AI (I assume we're mostly talking about LLMs) is useful for, and where it should be avoided, and many companies are currently making huge mistakes and trying to apply it in areas it is not suited to... Basically wishfu
between the hype and negativity (Score:2)
Yes (Score:2)
You should take every opportunity to learn a new tech, if you use it later on is another question, but you should at least have tried the basics.
Absolutely (Score:2)
And should management get a more realistic idea of what it is and what it isn't? Yep