AI naysayers who weren't around in 1994 when the web launched, don't realize they sound just like naysayers of the web when it was new. With pride, they resisted using it. "It's a time-waster," "just a tech fad," "can't find what you want," "AOL is all I need," they would all say. Don't be them.
I'm sure AI will have amazing accomplishments in many science areas, but I'm not sure that when AI starts charging what it needs to charge to cover the water and electricity bill that comes with 8 hour cybersex chat sesh that the price will be something people will pay.
"What can this web do for me?" That was another one heard a lot back then. (thanks for reminding me. :-) But when the question changed and became, "How do I use it to do things?" that's when it grew and changed the world. AI will follow the same trajectory.
The use-case is up to you. For example, my problem was deciding what to cook for dinner. I instructed AI to make an app to take photos of food in my fridge, analyze the items, and give me recipes to make. I didn't type one line of code. I only gave it instructions in plain English.
People didn't have to be forced to use the Web. People used it when they found a use for it. I wanted to talk to people about Mystery Science Theater 3000, so I went on Usenet. I wanted to share my comics online, so I made a webcomic. I wanted to learn about Totoro's penis, so I found fanfic sites.
What made the internet enjoyable, back in those palmy days, was that nobody gave a crap whether a venture capitalist somewhere thought it could make him a bunch of money.
I hear you. You just didn't hear me. It's not that AI is like the web, it's that the initial reaction to AI is similar to the initial reaction to the web. OK. I'm calm now.
even if it develops non-trivial use cases, you still need to solve for environmental concerns and ethical dilemmas, which feels increasingly unlikely given how AI dorks have had years to think about this and resort to denial
No. You're wrong. I'm a progressive who's utterly frustrated with the Dem party and its stance on technology. I mean, Al Gore, a Democrat senator, was the one who wrote the bill that funded the research that resulted in launching the Web. FFS.
Yes, the potential for unethical use of AI is a real problem. Scammers and hackers will swarm just like they have on the web. With the good, there will be bad. And just like with the web, the bad will be mitigated. It's a human endeavor. It'll never be perfect. But we shouldn't vilify it by default.
not potential. extant. genAI doesn’t exist without massive breaches of ethics. I don’t think you square that circle with the batch of vultures at the helm.
I created an app that allows you to take a photo of the items in your refrigerator, pantry, or freezer. You upload them, the AI model analyzes your photos, identifies the ingredients, and gives you recipes for the food items you have on hand. I didn't write a single line of code. I just had an idea.
In the time it takes you to faff about taking photos and uploading them I've already had a quick glance in the fridge and am three steps into making the meal, *and* I won't end up going hungry because an AI hallucinated and gave me an inedible dinner recipe.
Yeah, but I'm not as talented a cook as you are. I don't have that talent. So I found a way to use AI to solve my problem. But I think my example breaks people's thinking about AI and how it's only good for evil. Point is, it's evil to the extent that you use it to do evil things. ¯\_(ツ)_/¯
I'm not particularly talented, I'm just capable of basic adult functionality for the most part. Perhaps if you put your time into learning life skills rather than AI you too could achieve the basics without trashing the environment for it (which is evil)?
Thank you for asking so politely. But I must kindly refuse. If you don't agree with me, that's cool. But to request that I cease to express myself? Well now, that seems to be a sentiment originating from the other side of the political spectrum.
No silly. I don't need an LLM to tell me, I was there. The Internet has been around since the 60s. The HTTP protocol, which rides on the Internet, was developed by Tim Berners-Lee and released to the public in 1994. The first public websites started rolling out in mid 1994.
I was on AOL at 10/11 in the 90s so idk what you’re talking about lmao. I use several AI techniques for work and you know critical thinking allows me to approach with caution. My issue is with chat bots and people forming weird attachments to them. You’re typing shit that has nothing to do with me
Additionally, my feelings about ChatGPT having NOTHING TO DO with you. What is your investment in my feelings about it?? Stop taking things personally. My like or dislike is not an attack on your personality or identity. Get a grip.
I dunno, the web kind of accelerated ::waves hand:: all this so maybe those naysayers had a bit of a point. Plus, naysayers aren't ALWAYS incorrect about new technologies. In fact, a lot of new technologies don't go anywhere, but you don't hear much about them in the historical record.
I *was* around in 1994 and was using and could see the potential of the web. Even when the options actually available were limited the possibilities were very apparent to me.
I have used “AI” and it is handy for some limited use cases (https://goblin.tools “Magic ToDo” is simple and brilliant. The art plagiarism machines are evil but somewhat effective.).
I do not however see the wide horizons that the web implied, and AI is costly and insanely resource hungry to run.
You are spot on about the costs and resource hogging. But I hope prices will come down. As for the resource hogging, I'm working on a P2P GPU-as-a-Service platform where you contribute your idle GPU/CPU cycles to other users who are training/fine-tuning AI models. It's free open-source.
But they also sound just like the naysayers who were skeptical of theranos, and Google+, and NFTs and a million other tech fads that did not pan out. “Some people support it and some people don’t” is a pretty poor metric to decide whether something is a good idea.
That's a salient point. But time will tell. Personally, I'm fascinated by the possibilities. It's a new tech, some early use cases, such as the ubiquitous chatbots, will miss the target. But eventually that Napster-like killer use-case will land. And we'll forget about these conversations. :-)
“Left behind,” what a fucking joke. I wonder if they had meetings in the 60s to tell students to learn to use a calculator lest they get left behind in math. At least a calculator is probably going to return a correct answer…
Just the other day I saw that the "profession" of "prompt engineering" was dying out because AI is better at figuring out what questions to ask AI than any human ever could. And I'm just like, if you didn't see that coming you don't actually understand what AI is all about.
I must look into the Prompt Engineers Union, every librarian and graduate student and technical writer who ever built a database query should be a member automatically
Scott, you should be forewarned that to get into the Prompt Engineers Union requires you to go before the Prompt Engineering Board to receive your certification. Only properly certified Prompt Engineers are allowed to join the union. We can't have just anyone calling themselves a Prompt Engineer.
lol, picky but students weren’t using calculators in the ‘60s. I had to get one in ‘75 when I started college, and they were a pretty new thing, like nobody had them in high school. Maybe if you were rich?
Yeah, they were still on slide rules in the 60s. I remember playing with an LED calculator as a child in the 70s (making the word 80085 and such). In fact I borrowed a book from the library of games you could play with a calculator. I was an inside kid. :/
On two occasions I have been asked, — "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?"
I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.
From The Life of a Philosopher (1864) by Charles Babbage
This is actually a fantastic point: all my teachers in the 2000s told us NOT to use calculators "cause we wouldn't have one with us every where we went." And yet now it feels like every institution is saying "let the AI think for you, it'll be everywhere someday anyway, may as well get ahead of it."
It's a total 180 in just 15-20 years. And yet AI bros will use the calculator example as a reason WHY you should be using AI, totally missing the point behind why our teachers said that.
They weren't trying to predict the future, they were trying to make sure we understood what they were teaching.
Sometimes, anyway. My spouse teaches math and let a student retake a test recently because the kid's TI-80 was returning garbage on simple calculations, but the student didn't realize the problem until they had started the test.
I have been explicitly told that I have to use gen AI at work or else risk losing my job. So ya - I’m being forced. For now I am lying and saying “Claude helped me with this” when I just used my own fucking brain. But more and more my boss wants to see chat history, prompt crafting, etc.
This is also happening to me. Literally no complaints about the content of the deliverables or how long it is taking me to do it - management wants to know that i used AI for it, for some fucking insane reason. So I lie.
It’s heartbreaking because I am really good at my job. I was originally hired because I have years of experience and knowledge and highly developed problem solving and critical thinking skills. Why would I want to hand over all the value I bring to a word prediction machine?
It's also possible they're trying to build a library of chats and prompts to figure out whether they can replace most of the employees with a few AI promoters and proofreaders.
Because the company spent money on AI vendors that werent fit for purpose and the people who impulse-grabbed at the AI feel the need to justify their bad decision
It’s driving all the layoffs in tech, too. AI is actually more expensive than just hiring smart people, but the c suite thinks we need to be all in on AI and they gotta find the budget somewhere.
Funny you should phrase it like that. A synonym is in fact "Machine Learning."
Here's a handy guide by an academic I follow who hates the many misuses of the term and incorrect claims about AI. The table is for the various ways the term AI can be used and synonyms.
I can’t overstate: these idiots threw money at crypto, blockchain, NFTs, AR/VR - AI really is their white whale after all the prior gizmos had few $$$ commercial uses.
They know consumers won’t voluntarily buy enough AI to make their investment back, so they’re capturing the public purse.
You’d think, but apparently people don’t know what they like and need to be educated. Frequent refrain: “no one would have asked for an iPhone!” (this is an updated version of car/horse)
Now they make stuff they think will make money. Then try to entrap people. Horrible experience ensues.
They spent 20 years telling people art was for losers! They’ve sunk too much into the idea that Everyone Should Be STEM And Art Is Worthless. Can’t admit they were wrong now.
We’re still here, they’ve just convinced themselves they can understand motivation, habit, behavior through analyzing data. And maybe send a new hire BA to ask three people leading questions. Oh and “everyone on the product team is a researcher!”
They'll never make the investment back because there's no economy of scale. ChatGPT loses money on paying customers and 2 X and many queries costs 2 X as much. Unless it fundamentally changes, it will keep losing money. They want more users to pull in more investment capital.
The game’s changed. Now they want to dismantle enough of the government using interns that they can say “oh dear, they meant well - since it’s a national emergency to get back up and running, have our AI solution for cheap”.
And then the govt service contracts and subscriptions are locked in.
The stupid thing is - yes, there are valid use cases for AI! But specialized cases that don’t make techlords oodles of cash, like medical imaging comparison, or specific use cases like specialists using ChatGPT with very targeted queries.
Those specialized cases also require very high levels of accuracy, which means the input has to be validated by experts, not just sucked up from publicly available data. So they cost a lot to make AND have limited markets, the opposite of what they want.
Ding ding. So the whole “you don’t need an expert, you just write a prompt and ChatGPT is your expert!” is immediate nonsense.
Experts need to input the data, experts write the precise prompt, experts need to validate. The experts are the golden geese here. AI just accelerates egg production.
An inability properly to grasp nuance is something that's quite noticeable about this generation of chat bots, and I suspect a level of comprehension will always elude them, no matter how superficially advanced they get. You have to dumb everything down. It's like talking to a child, or an American.
No but I was forced to attend an all day lecture on using Chat and other GAI in my work, and proselytized to about getting left behind. I used the time to passive aggressively write a 12 paged essay how it was going to get us all killed.
I had to take a course on using it. The interactive lab assignment was to get it to make a simple list of how to start a dog walking business. And the course tells you to try not to get too frustrated if you can't get it to even do that.
Ok you got me. It was 12 pages of handwritten notes outlining my thoughts for an academic article, so technically a rough outline. All I had to work on was my Boox tablet.
The day culminated in group brainstorm sessions, after which I realized that my manager's way of 'processing all our suggestions' likely meant feeding them into ChatGPT
I have yet to get the answer I need from a chatbot… ever! “Sorry, I didn’t understand your question” is what I usually get! Or generic stuff that I already tried!
Well to be fair, traditional chatbots don’t use GenAI but rather a library of “intents”. This is in some ways good, as it ensures only vetted information is shared, but it’s frustrating when you ask it something it doesn’t have an “intent” for. GenAI purports to solve that, but can’t.
Comments
It's like Google search.
It's a tool, not a new way of life.
Jesus Christ calm down.
1/
I do not however see the wide horizons that the web implied, and AI is costly and insanely resource hungry to run.
2/2
I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question.
From The Life of a Philosopher (1864) by Charles Babbage
They weren't trying to predict the future, they were trying to make sure we understood what they were teaching.
Still no idea what happened to it.
It’s driving all the layoffs in tech, too. AI is actually more expensive than just hiring smart people, but the c suite thinks we need to be all in on AI and they gotta find the budget somewhere.
Or as Blueprint said:
“Learning Machine” wouldn’t sound as impressive as “Artificial Intelligence.” It’s deceptive marketing really.
Here's a handy guide by an academic I follow who hates the many misuses of the term and incorrect claims about AI. The table is for the various ways the term AI can be used and synonyms.
https://bsky.app/profile/irisvanrooij.bsky.social/post/3k4bptpv3tf2j
They poured tons of money into AI, they tried to sell it commercially, they jammed it into every possible service and people STILL said “nah”.
So they’re destroying the government in order to MAKE you fund/use it.
They know consumers won’t voluntarily buy enough AI to make their investment back, so they’re capturing the public purse.
Now they make stuff they think will make money. Then try to entrap people. Horrible experience ensues.
UX has been a bloodbath over the last five years.
And then the govt service contracts and subscriptions are locked in.
It’s cool tech. But it’s not a golden goose.
Answer: “…it’s CUTTING EDGE!”
My transparently negative expression is probably why I got rolled off that project.
Experts need to input the data, experts write the precise prompt, experts need to validate. The experts are the golden geese here. AI just accelerates egg production.
Who needs human accomplishments anyway?