For OpenAI, o1 represents a step toward its broader goal of human-like artificial intelligence. More practically, it does a better job at writing code and solving multistep problems than previous models. But it’s also more expensive and slower to use than GPT-4o. OpenAI is calling this release of o1 a “preview” to emphasize how nascent it is.
The training behind o1 is fundamentally different from its predecessors, OpenAI’s research lead, Jerry Tworek, tells me, though the company is being vague about the exact details. He says o1 “has been trained using a completely new optimization algorithm and a new training dataset specifically tailored for it.”
OpenAI taught previous GPT models to mimic patterns from its training data. With o1, it trained the model to solve problems on its own using a technique known as reinforcement learning, which teaches the system through rewards and penalties. It then uses a “chain of thought” to process queries, similarly to how humans process problems by going through them step-by-step.
At the same time, o1 is not as capable as GPT-4o in a lot of areas. It doesn’t do as well on factual knowledge about the world. It also doesn’t have the ability to browse the web or process files and images. Still, the company believes it represents a brand-new class of capabilities. It was named o1 to indicate “resetting the counter back to 1.”
I think this is the most important part (emphasis mine):
As a result of this new training methodology, OpenAI says the model should be more accurate. “We have noticed that this model hallucinates less,” Tworek says. But the problem still persists. “We can’t say we solved hallucinations.”
It’s irresponsible because making it sound like it’s true AI when it’s not is going to make it difficult to pull the plug when things go wrong and you’ll have the debate of whether it’s sentient or not and if it’s humane to kill it like a pet or a criminal. It’s more akin to using rainbow tables to help crack passwords and claiming your software is super powerful when in reality it’s nothing without the tables. (Very very rudimentary example that’s not supposed to be taken at face value).
It’s dangerous because talking about AI like it’s a reasoning/thinking thing is just not true, and we’re already seeing the big AI overlords try to justify how they created it with copyrighted material, which means the arguments over copyrighted material are being made and we’ll soon see those companies claim that it’s no different than a child looking up something on Google. It’s irresponsible because it screws over creative people and copyright holders that genuinely made a product or piece of art or book or something in their own free time and now it’s been ripped away to be used to create something else that will eventually push those copyright holders out.
The AI market is moving faster than the world is capable of keeping up with it, and that is a dangerous precedent to set for the future of this market. And for the record I don’t think we’re dealing with early generations of skynet or anything like that, we’re dealing with tools that have the capability to create economical collapse on a scale we’ve never seen, and if we don’t lay the ground rules now, then we will be in trouble.
Edit: A great example of this is https://v0.dev/chat it has the potential to put front end developers out of work and jobless. It’s simple now but give it time and it has the potential to create a frontend that rivals the best UX designs if the prompt is right.
I appreciate the effortful response but i dont think regulators would get caught up on colloquial names when weighing benefit versus harm and deciding to do something like ban a model.
We just arent close enough to the same perspective to discuss it further. Thanks again for the good faith clarification.
I think over-selling the “AI” with “reasoning/thinking” language becomes fraudulent and encourages inappropriate/dangerous applications.
Why does ai that has a “reasoning” step become dangerous?
It will be used to take control over peoples lives.
In any simple way it may be - denying job/insurance/care/etc, it will be hailed as using ‘reason’, while it just repeats patterns from the training sets.
It does not ‘reason’, because it can’t. Trying to sell it as such is very dangerous as it will be used against people, and it’s dishonest for the investors as well, as they will jump on it even though it’s not ‘true’ and it never will be for this model.
https://softwarecrisis.dev/letters/llmentalist/
I agree that is a bit of an ethical minefield to employ it to make decisions that affect peoples livelihood. But my point is if a company uses it to decide if an insurance claim should be paid out, the models ability to make those decisions isnt changed by what we call the steps it takes to come to a decision.
If an insurance company can dissect any particular claim decision and agree with each step the model took, then is it really different than having someone do it? Might it be better in some ways? A real concern is the fact that ai isnt perfect and mistakes made are pretty hard to accept… seems pretty dystopian i get that. But if less mistakes are made and you can still appeal decisions then maybe its overblown?
LLMs just repeat training sets - so every mistake is repeated forever.
Every bias is locked in and can’t be fixed.
So you just deny people and expect them to appeal everything… sounds like you are offloading costs on the victims.
Shit like that is what makes it demonic.
I think its the company’s responsibility to incorporate a technology to carry out their policy accurately. They cant just use an LLM stock from a vendor. They work to adapt it for their needs and get acceptable results. I think if an llm isnt considerably more accurate than humans then its a disservice to their customers and they should be responsible for that. There should be regulations to keep companies from using models if they dont work
assuming that “AI” has “reasoning” and using it in applications that require that is dangerous
Sorry but thats not an explanation of your position, thats restating what you just said.