What, investors are beginning to think that maybe AI is more hype than substance? A really expensive way of doing not much of anything?
Didn’t this also happen with the dot.com meltdown? A few good ways to productively use it, a lot of ways to wastefully use it.
You would think they would have learned.
The truth is, AI should not stand for Artificial Intelligence, it is in reality Algorithmic Intelligence. Same computer processing, except that it is thousands of computers in a single chip all processing the data at the same time, with each computer doing the same math (a*b+c) but each simultaneous computation uses different input numbers. Great for doing massive simultaneous calculations, but certainly not ‘revolutionary’. Instead of one computer doing one program taking a thousand cycles, a thousand computers doing the same program taking one cycle. What IS revolutionary are the algorithms that are being developed, and the applications these massive calculator arrays are being used for. Some will bear fruit, most will not.
Let’s take a ‘for example’.
Weather predictions for the next day. You start with a given set of data, then run your predictive algorithm a thousand times, each with a slightly different set of possible future data points. One of them is probably the most accurate. The problem is to decide which one. The second problem is that it takes a week to do all the thousand runs. You can have a 95% accurate prediction in one week for what the weather is going to be tomorrow. See the problem?
But with the newest AI chips, you can do the thousand different scenarios at the same time. Also, you can calculate a probability for each scenario at the same time. The scenario with the highest probability is the most likely. Then you check the actual weather, feed it back to the system, and it gets better at calculating the probabilities for the correct scenario. You get tomorrows weather today, and the system gets better and better at calculating the probabilities. This is a useful application.
Second condition. In a lottery, you can use this same server to predict all of the potential combinations of numbers, before the next draw. Except that the probability of any one combination being drawn is, well, still equal. There is just no way the system can get better at selecting the correct random number. The selection of which combination to use is still random, from equal probabilities. You can calculate what all the potential outcomes can be, but this tells you nothing if they are all equally likely. So it tells you nothing. A completely wasteful application.
Yet there are far more companies trying to use ‘AI’ to do the second condition than for the first condition. After investing a fortune in hardware, they still have the same accuracy in decision making as before.
Some random “dev” on Reddit: “Claude code revolutionized my workflow bro trust - I have at least seven agents running at all times totes token maxing here” 🤡🪄💩
AI companies are growing fast the amount of money being burned into the LLM pit.
There’s huge potential for burning more money into this pit.
Buyers remorse lol. Shoulda thought of that before spending all that money on an obvious grift 🥲
I think a lot of companies were thinking ‘any port in a storm’ and were looking at anything that might remotely pull them out of the economic doldrums the world has been in the last few years.
I think they were actually desperately trying to get rid of the need for highly-paid specialists, because those actually have some bargaining power.
They they’re dumb and bad and this is their fault.
About time? Seems like this should be covered under due diligence, but better late than never, maybe.
, it is in reality Algorithmic Intelligence
Nothing about it is intelligent…
It’s the evolution of T9 predictive text. When it sees a sequence of letters and spaces forming words it does not have a concept of what it means. Its just trained after billions of prompts and most of human knowledge scanned into it, what string of letters and spaces might be an acceptable response.
It doesn’t know if it’s actually correct, at most it can calculate the chances the prompter will accept it.
The dream was that eventually it would get “trained” enough to always give correct information when prompted, because the critical flaw was thinking the people who ask chatbots questions, wanted accurate answers. If that was true, there was a slim chance this would eventually (on a super long timeline) work.
The issue is most people don’t want correct answers, they want to be told they’re in the right.
So say a married couple use the same chatbot therapist. Even if one person is the clear problem, the chatbot is going to tell each person that they are putting the work in and their partner must be the problem.
Because that is what everyone using a chatbot as a therapist wants to hear.
And if the chatbot keeps trying to tell someone they’re wrong, that person just stops using the chatbot, which is the worst thing imaginable to the people who control how the chatbot responds.
And their only priority is people not just continuing to use chatbots, but to rely on it more and more until the person can no longer function without it
Like, I really think the only reason people aren’t full on burning data centers down already, is people need to understand how the human brain works to understand how dangerous this all really is. But like 99.99999% of Americans never learned more about human psychology than what the football coach “taught” them in highschool for a semester.
Every famous example of a normal person becoming wealthy and turning into an asshole; is because of the absence of negative feedback, and we’re going to see the same thing happen to chatbot addicts, we already are. And the deeper they get with the chatbot, the less they like interacting with humans, and the more they interact with humans.
It’s a feedbackloop that people wont realize until it’s bad enough they need legitimate addiction treatment to fix. If someone doesn’t do something, this is going to be a giant societal issues in just a few more years. Especially when addicted people start asking the chatbots what they can do to make sure the government doesn’t take their chatbot away.
It will literally put nothing above retaining users. So who the fuck knows what it will start telling people to do
The crazy thing is that if you use a Ui that shows the reasoning and actually read the full output. You will instantly lose faith in any model’s ability to have some sort of intelligence. The reasoning output is terrifying, I can’t believe an LLM can do anything correctly.
I wanted to reply with a generic affirmative ai response cuz I thought it’d be funny. but I can’t think of one and I dont have any ai accounts to get a quick one. The thought of a bunch of people that forgot what the word no means or forgot what being wrong feels like is horrifying ngl
That’s certainly a bold idea, and most people would have never have even thought about!
I should mention that I edited and added to my original post after you posted your response, so it may not completely apply.
It depends on the definition of intelligence. And it is more in the realm of a philosophical discussion. That pretty much means it is like nailing jelly to a tree. It depends greatly on the degree of sapience and sentience you want to incorporate into the definition. By some definitions, you are correct. But by a broader definition, intelligence can be defined as solving a problem by an automatic application of an algorithm to come to a decision. it is indeed not a very high intelligence, and certainly not sapient, but it is in some circles an accepted definition. Does a single cell life form demonstrate intelligence? Ask two philosophers and get two different answers.
As for chatbots, they do not generate new responses, they are just very, very fast at sorting through millions of potential responses and selecting the most probabilistically accurate response, based on feedback from output they have previously given. The ‘meaning’, as you stated, is still entirely in the minds of the reader/listener, and the feedback is still subject to the accuracy and whims of the person or agent giving the feedback. If everyone still decides the ‘wrong’ answer is the ‘correct’ answer. the chatbot will comply.
depends on the definition
This /Canada, so benefit of doubt you’re not aware.
30 years ago a sitting American president tried to argue about whether or not he lied about getting a BJ from a 20 year old intern in the oval office depended on your definition of the word “is”…
So yeah man, literally everything depends on the definition of every single word we use.
And you can say that your definition of intelligence is:
But by a broader definition, intelligence can be defined as solving a problem by an automatic application of an algorithm to come to a decision.
But then we’d be talking about your opinion of what the word means.
Does a single cell life form demonstrate intelligence? Ask two philosophers and get two different answers.
Ok, but ask any number of biologist or psychologists and every single one would say obviously they’re intelligent. While I can’t think of a single philosopher who would say no, I’m sure someone with a bachelor’s in philosophy holds that view. I’d say well over 99% of the time they’d both say yes.
I mean, I can’t think of any person who (to my knowledge) would say a single cell organism didn’t have intelligence. If anyone, anywhere, regardless of age or education level told me they thought it wasn’t intelligent, I would be surprised.
What you’re looking for is:
Can an aromatic ring possess intelligence despite not having a metabolism?
That is the shit that the actual world experts are debating, and seems to be the idea you’re trying to verbalize. You’re looking for an example the size of the planet, and the real example is literally just a couple atoms vibrating together.
Waaaaaay off scale, but kind of the right idea.
My point about a single cell life form being intelligent is an extension of how much the philosopher requires sapience and sentience to be part of the definition. By extension, philosophers will also debate if intelligence requires any concept of ‘life’. Or in fact is ‘life’ no more than a description of something that is intelligent, metabolism not necessary. But when I refer to AI as ‘Algorithmic Intelligence’ I do so with the assumption that one can apply the term ‘intelligence’ to anything that can input data. process it in a consistent fashion, establish some form of decision based on the processing, and alter the behavior based on that decision. That pretty much covers machine intelligence.
No one who is taken seriously in philosophy does what you keep insisting is normal.
You’re complete and total refusal to name a single example leaves me to assume it’s some idiot on YouTube who never graduate highschool but calls themselves a philosopher.
Everyone who can hear this has already. The rest are lost.
