Most people here have been using AI in some form for their entire lives without knowing it. It just did its job quietly with nobody noticing. Then venture capital (or just capitalism itself) ruined everything and broke the contract: publicly acquired data must be given back to the public for free.
I could pontificate at length about the terminology and how it has gotten fucked. The blending of the terms itself is part of what makes it difficult to have a reasonable and nuanced discussion.
Let’s take a moment to separate out AI from machine learning from deep learning from LLMs.
AI is fucking old. It used to mean “any algorithms that create intelligent behavior”. Not a particularly useful definition these days, but it used to mean things like pathfinding and searching.
Machine learning is a more useful phrase: a set of algorithms to solve problem where we don’t know “how”, but we have examples of inputs and outputs. For example, I don’t know how I would define cute, but if someone showed me a bunch of photos I could probably say which ones were cute, not cute, and unsure.
Deep learning is a subset of machine learning that uses a specific set of algorithms: neutral networks.
LLMs are a subset of deep learning that use “transformers”. Which is a specific architecture that does a lot of things quite well, like determine how proteins fold, how drugs interact, how words interact in a sentence, etc.
If you’ve used Google Maps at any point since it was created, you’ve used classical AI.
If you’ve used email, you’ve used machine learning.
If you’ve used a photos app that lets you search for similar pictures of people, you’ve used machine learning.
If you’ve had more than one prescription filled in the past five years, your pharmacist has used AI (even if they don’t know it) to check potential drug interactions.
Don’t get me wrong, I fucking hate that the field I spent my whole life researching has been coopted into a way to siphon money from people into the coffers of the richest fucking parasites, but when people say “fuck AI” they have either lost the nuance or never had it. Everyone that hears the message experiences on the surface and it does them a disservice.
When the luddites broke the textile looms, did they hate the machines or did they hate the loss of their livelihoods?
When the early industialists broke into factories and smashed their equipment, did they hate the machines or did they hate the captains of industry that forced them to work inhumane hours in terrible conditions?
When people say “fuck AI” do they hate the math that, until this point, has led to a better world for us all, or do they hate the system that has enshittified it into one of pure exploitation?
This whole mess feels like a distraction to me. Tech should be a social good. It should be helping people. Not to say it’s without problems, but now when we say “fuck AI” it leads us to pushing back against technology itself rather than the system that’s using it to hurt people.
Most people here have been using AI in some form for their entire lives without knowing it. It just did its job quietly with nobody noticing. Then venture capital (or just capitalism itself) ruined everything and broke the contract: publicly acquired data must be given back to the public for free.
I could pontificate at length about the terminology and how it has gotten fucked. The blending of the terms itself is part of what makes it difficult to have a reasonable and nuanced discussion.
Let’s take a moment to separate out AI from machine learning from deep learning from LLMs.
AI is fucking old. It used to mean “any algorithms that create intelligent behavior”. Not a particularly useful definition these days, but it used to mean things like pathfinding and searching.
Machine learning is a more useful phrase: a set of algorithms to solve problem where we don’t know “how”, but we have examples of inputs and outputs. For example, I don’t know how I would define cute, but if someone showed me a bunch of photos I could probably say which ones were cute, not cute, and unsure.
Deep learning is a subset of machine learning that uses a specific set of algorithms: neutral networks.
LLMs are a subset of deep learning that use “transformers”. Which is a specific architecture that does a lot of things quite well, like determine how proteins fold, how drugs interact, how words interact in a sentence, etc.
If you’ve used Google Maps at any point since it was created, you’ve used classical AI.
If you’ve used email, you’ve used machine learning.
If you’ve used a photos app that lets you search for similar pictures of people, you’ve used machine learning.
If you’ve had more than one prescription filled in the past five years, your pharmacist has used AI (even if they don’t know it) to check potential drug interactions.
Don’t get me wrong, I fucking hate that the field I spent my whole life researching has been coopted into a way to siphon money from people into the coffers of the richest fucking parasites, but when people say “fuck AI” they have either lost the nuance or never had it. Everyone that hears the message experiences on the surface and it does them a disservice.
When the luddites broke the textile looms, did they hate the machines or did they hate the loss of their livelihoods?
When the early industialists broke into factories and smashed their equipment, did they hate the machines or did they hate the captains of industry that forced them to work inhumane hours in terrible conditions?
When people say “fuck AI” do they hate the math that, until this point, has led to a better world for us all, or do they hate the system that has enshittified it into one of pure exploitation?
This whole mess feels like a distraction to me. Tech should be a social good. It should be helping people. Not to say it’s without problems, but now when we say “fuck AI” it leads us to pushing back against technology itself rather than the system that’s using it to hurt people.