AI is absolutely damaging the job market, largely because executives are desperate to pretend it can do things it can’t, but I would feel more apprehensive about apocalyptic predictions of the end of all white collar work if they weren’t all coming from people trying to sell chatbots
I have found it useful for doing boiler plate stuff that I could figure out but don't have time/feel like doing. Has saved several hours over several taks. Having said that, I would not pay $200 a month for it, maybe $10-20. However, I can see it being worth it for large employers to pay for it.
I feel like this could be accomplished for free by the use of actual boilerplate like we used to
I know people hate AI on bluesky, and yes it is over hyped, but it is also genuinely useful to a lot of people. It is providing value now. People can feel free to not use it, but it isn't going away
If you want to justify using the unethical plagiarism machine that destroys the environment to do something humans managed for decades to do with other tools without that baggage, don’t bother
The environmental footprint of agriculture is like 10 orders of magnitude greater than data centers. Stochastic gradient descent is not "plagiarism", and the data rights issue can be solved by forcing companies to pay for training data. But if you want to be ignorant of how it works, you do you
Something can be very useful to some and still a net negative for society. The companies funding and building these tools, along with the underlying business model, cannot work with limited adoption so it's being forced on all of us and causing societal problems
Companies are shoving it into everything, but I have not noticed a price increase at the end user level, and I have been able to turn it off/avoid it as I want 🤷. Have no clue at what price it would be profitable though. It ain't worth $200/month to me. TBD if business subscriptions are profitable.
When a software company adds GenAI features, it's at the expense of other features or bug fixes. Shoving it everywhere means my payroll software is ignoring bugs for AI. And Microsoft just didn't tell their customers they would all start paying for AI
As for profitability, what i mean is that companies have invested over a trillion on this technology. They cannot make a profit but some people using it for personal use, they need enterprise wide use. So there is no future where you get some AI and I can ignore it. Its all or nothing
At some point, they need the cost of inference to either come down greatly or for people to pay more for inference. The latter ain't happening, and I don't see how shoving AI into software without increasing the price solves this. It will catch up to them if they are that far over their skis.
Yes, their strategy is get GenAI plugged into your business so that when they increase the price, you have to accept it. Which is part of why this tech is an existential crisis, even if some people are finding benefits.
I guess, in my experience it is not seen as some widespread irreplaceable super assistant. Usage is more targeted than that. These things aren't AGI, they are a more intuitive search interface that have the side effect of occasionally bullshitting. If costs get excessive, usage will drop
Most don't realize our entire economy has been slowly pivoting to be focused on GenAI and LLMs
The sales talk is working in executives because: 1) They don’t actually know how to do any of the things AI purports to do. They see it output code/design/text, but they don’t have the expertise to judge the quality of that output. 2) It gives them an excuse to “reduce overhead”. Line goes up.
What I'm really afraid of is how many jobs might settle in the unhappy local "optimum" where AI is used in spite of not being up to the task and then a human being has to babysit it and correct it. This sucks and I suspect is not actually more productive, but it's what is happening to translators.
This and also the resulting enshittification of everything. They'll fire all the humans, put certain tasks in the hands of AI, and if it becomes unusable, welp, that's someone else's problem and now unrelated to whether they should have fired all the humans.
They'll hire back humans at half the rate because "the ai does most of the job"
I've seen the opposite too: the fired humans creating consultancy services in solving AI messes and charging triple their previous rate.
Reports I'm seeing indicate it's impacting the junior developer job market but not the senior developer job market. None of them seem to address the question of where senior developers come from. I'm waiting for the thinkpieces in 5 or 10 years pondering why there aren't enough senior developers.
Yeah, that’s my biggest worry. I think it’s going to be 10/15 years, but there’s a whole generation of employees that either never learned how to do things, or didn’t get hired at all.
From the heart of the genocide in Gaza, from beneath the rubble, from the jaws of starvation and the darkness of oppression… I cry out — not to surrender, but to search for a thread of hope that might lead me back to life. ♦️Please, help me… and let your light touch my heart once more💔
The number one application of AI seems to be doing high school and undergrad homework at varying levels of success
Which then sells people with a high school to undergrad level of understanding of things on the idea that AI is "a supergenius". It can scrape Wikipedia, Brian.
I feel the same way about AI as I do about incarcerating someone at $200K/yr rather than educating and feeding them adequately as kids We could just do the cheaper thing and invest in an array of mediums with money distributed by people on the field like the NIH But mid people really hate artists
seriously, every time you read a “nobody will have a job in ten years, it’ll all be AI” prediction it’s coming from Sam Altman or some AI agent startup exec trying to sell you on the damn things
It's the worst selling point I've ever heard. "It'll put you out of a job!" And then how will we pay our bills, Sammy? How will economics work when money goes nowhere, Sam? Hubert Turvy could make a smarter choice ffs
It's frustrating how hard it is to have proper conversations about it over the din created by all the hype! My theory is the promoters think - and they may be right - that the more noise there is, the more it looks to investors like it's a real thing.
Sam Altman is an AI character created by jkbowling
Similar to how the "70% of Americans live paycheck to paycheck" only ever comes from companies selling financial products or advice to scared people
esp. since the obvious candidates to be replaced by LLMs are executives and high up managers, as the cost savings would be enormous
We’ve passed the point where any given AI take says far more about you, what you do, and your vision of society than anything about the technology itself. Only thing that was comparable like this is crypto. Even the internet wasn’t quite so…sociological.
I was just in India for business and one of my new teammates there was showing off our new AI coding assistant. He right-clicked on a variable name to try and figure out how many times it appeared in the code. The tool said one. I said "It's at least three." It was, indeed, three.
It reduces the need for lots of entry level roles where basically they did grunt work to start familiarising with a field but AI does it instead
Hey now! Sometimes it's coming from 'reporters' whose entire beat is hyping up AI!
The people selling/marketing it don't even know what most work looks like but they've infected the world with this idea that better technology will make our jobs easier when each new wave of technology has just made our work more complex
AI is a previously worn, pus filled bandaid, made for bosses with Dark Triad traits that hate having to directly deal with people
One thing I’ve been thinking about lately is how good domain tuned LLMs could be vs the generalized all purpose systems being put out at the moment.
I was genuinely told today at work to use an LLM to tech edit a deliverable. Everyone on the team collectively looked at each other with visceral disgust that they were making us do it.
This has been one of my biggest gripes with AI, as a software engineer. People are insisting on foisting AI into problem spaces and creating new problems rather than providing solutions. All at the expense of a gazillion other ethical and ecological concerns.