505
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
this post was submitted on 17 Jul 2023
505 points (93.2% liked)
Technology
59414 readers
1100 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
Part of the problem with AI is that it requires significant skill to understand where AI goes wrong.
As a basic example, get a language model like ChatGPT to edit writing. It can go very wrong, removing the wrong words, changing the tone, and making mistakes that an unlearned person does not understand. I’ve had foreign students use AI to write letters or responses and often the tone is all off. That’s one thing but the student doesn’t understand that they’ve written a weird letter. Same goes with grammar checking.
This sets up a dangerous scenario where, to diagnose the results, you need to already have a deep understanding. This is in contrast to non-AI language checkers that are simpler to understand.
Moreover as you can imagine the danger is that the people who are making decisions about hiring and restructuring may not understand this issue.
The good news is this means many of the jobs AI is "taking" will probably come back when people realize it isn't actually as good as the hype implied
Not quite. It's more that a job that once had 5-10 people and perhaps an "expert" supervisor will just be whittled down to the expert. Similarly, factories used to employ hundreds and a handful of supervisors to produce a widget. Now, they can employ a couple of supervisors and a handful of robot technicians to produce more widgets.
The problem is, where do those experts come from? Expertise is earned through experience, and if all the entry-level jobs go away then eventually you'll run out of experts.
Education. If education was free this wouldn't be a problem, you could take a few more years at university to gain that experience instead of working in a junior role.
This is the problem with capitalism, if you take too much without giving back, eventually there's nothing left to take.
You don't get experts from education. You get experts from job experience (after education).
You definitely don't get experts from unemployed people, or from people working to the bone doing menial labor for minimum wage.
Education is a broad term, that could include apprenticeships where you do get real work experience. And education would have to change a lot in all areas. The point is, the government can support people to gain that experience, the problem is that right now it isn't. It's common to exit just a bachelors degree with crippling amounts of debt.
And it's viewed more positively in the society to have a bullshit Bs or Ms than a (usefull) trade degree
I wasn't commenting on what type of education is better or worse than another. The point is that we need to support people through education.
It’s just that I fear that realisation may not filter down.
You honestly see it a lot in industry. Companies pay $$$ for things that don’t really produce results. Or what they consider to be “results” changes. There are plenty of examples of lowering standards and lowering quality in virtually every industry. The idea that people will realise the trap of AI and reverse is not something I’m enthusiastic about.
In many ways AI is like pseudoscience. It’s a black box. Things like machine learning don’t tell you “why” it works. It’s just a black box. ChatGPT is just linear regression on language models.
So the claim that “good science” prevails is patently false. We live in the era of progressive scientific education and yet everywhere we go there is distrust in science, scientific method, critical thinking, etc.
Do people really think that the average Joe is going to “wake up” to the limitations of AI? I fear not.
And AI is not always the best solution. One of my tasks at my job is to respond to website reviews. There is a button I can push that will generate an AI review. I've tested it. It works... but it's not personal. My responses directly address things they say, especially if they have issues. Their responses are things like, "thanks for your five-star review! We really appreciate it, blah blah blah." Like a full paragraph of boilerplate bullshit that never feels like the review is addressed.
You would think responding to reviews properly would be a very basic function an AI could do as well as a human, but at present, no way.
This assumes that your company doesn’t decide the AI responses are good enough in exchange for the cost savings of removing a person from the role, and that they don’t improve in a subsequent update.
True, although my company emphasizes human contact with customers. We really go out of our way with tech support and such. That said, I hate responding to reviews. I kind of wish it was good enough to just press the 'respond to review with AI' button.
This.
In accounting, 10 years ago, a huge part of the job was categorising bank transactions according to their description.
Now AI can kinda do it, but even providers that would have many billions of transactions to use as training data have a very high error rate.
It's very difficult for a junior to look at the output and identify which ones are likely to be incorrect.