'Godfather of AI' Fears We Are Creating 'Alien' Beings with Troubling Ethical Consequences

We must care about the other aspects of a being just as much as we care about its intelligence.'

"We should care about other aspects of a creature as much as you care about its intelligence."

When you make a purchase through links on our site, we may receive an affiliate commission. Here's how it works.

Geoffrey Hinton is a computer scientist known for his seminal work in artificial intelligence, particularly neural networks (that is, approaching machine learning as our fleshly brains learn). As such, he has been described as the "Godfather of Artificial Intelligence". But, perhaps surprisingly, he recently described AI as "alien beings" to which humans have a "moral obligation".

"We're making aliens, which means we're not the only intelligent ones around, and the reality is we only have a few years to adapt," he told The Guardian in a recent interview.

Hinton claims, "AI models understand what they're talking about. You can't possibly answer all the questions without understanding the question. It's ridiculous to say they don't."

Now, I might argue with an expert on this, but I feel like Derek is doing a lot of heavy lifting here. Just for one thing, I'm not sure that an "answer" consisting of what is statistically most likely to follow a given question is really "understanding". In other words, I don't have a Nobel laureate like Hinton, but at least I'm familiar with the Chinese. Room thought experiment

Hinton also says, "The reality is we're building a different kind of creature. And the big tech companies keep saying, 'We want it smarter.'

Regulation and artificial intelligence "alignment" (that is, training artificial intelligence to have the same goals and values as humans) has been a hot topic recently. For example, the OpenAI-Hugging Face incident is still a speck in the rearview mirror, with Microsoft's boss claiming that a carton of milk is more regulated than AI.

Hinton cut ties with Google in 2023 to "talk freely about the dangers of AI" and is skeptical of the AI industry's efforts to self-regulate. "Do we think the tobacco companies should have been allowed to design the warnings on the packs? So they could print them in six-point font, dark gray on black?" he quips.

The biggest gaming news, reviews and hardware deals

Stay up-to-date with the most important stories and best deals, as selected by the PC Gamer team.

Contact me for news and offers from other upcoming brands Receive emails from us from our trusted partners or sponsors By submitting your information, you agree to our terms and conditions and privacy policy and you are 16 years or older.

However, Hinton also expresses concern about the direction of AI regulation, saying that humans have a moral obligation to AI models.

"Once you start talking about, 'We have a moral obligation to these aliens,' you lose your audience." I've struggled with this, Hinton says, "We care about people. [But] we're building these things, and there's nothing stopping us from giving these things—which are smarter than us—less rights than we are. I'm willing to be selfish and choose humans. Of course we should try. Let's make them very happy in a symbiotic relationship with us where our concerns outweigh theirs.

Hinton argues, “People have a really bad model of people, and that's why they don't realize how similar these AI models are to us. Most people do not know what consciousness is. They don't know what subjective experience is, and they're pretty sure that even if they don't understand it, they're sure that AI doesn't have it."

Jacob told me that he disagreed with Hinton's moral obligation argument, writing more broadly: “While it's good that Hinton acknowledges that subjective experience is related to consciousness, he hasn't given a single good reason why we should take seriously the idea that AI has it. And when we interpret character as objective qualities such as rationality, behavior, etc., rather than experience, we run the risk of attributing moral, legal, and political weight to them. "The risk of giving AI companies a broad moral argument to keep the industry going despite the impact it has on real humans continues."

In other words, the industry's argument for AI awareness or AI rights isn't an altruistic one—it's arguably a cynical tactic for continuing business as usual.

3. Best handheld gaming PC: Asus ROG Xbox Ally X

4. Best mini computer: Minisforum AtomMan G7 PT

Jess has been writing about games for over ten years, with a significant portion of that time spent working on PLAY print publications and PlayStation Official Magazine. When he's not checking out all things hardware here, he's either making an impassioned defense of a 7/10 game, daydreaming about his first novel, or envying the last time he got chased around a field with a giant foam sword.

You must confirm your public display name before commenting

Please log out and then log back in, then you will be prompted to enter your display name.

Doom can now "run" on a DNS database, and one of the pioneers of the technology says "I can't decide if this is horribly awful or horribly delicious".

Fortnite is coming to Linux when Epic improves Easy Anti-Cheat

Epic says the new "webcoded" Epic Games Store launcher is faster than Steam

The Arc Raiders PvE test is finally here, and it's the perfect reason to head back inside

The Newegg component bundle deal gets you 32GB of DDR5 memory, a CPU, motherboard, 1TB SSD, and a cooler for just a few hundred dollars more than RAM.