Everyone Is Wrong About AI Except Me | Hacker News Reader

Everyone Is Wrong About AI Except Me

jasonpargin.substack.com

32 pointsteddyh31 comments

Interesting first point. I'm reminded of Blade Runner, which definitively fits in the "robots are people too, you monster!" genre ... But the source material was the exact opposite. Philip K. Dick went out of the way to rub in that even if the robots could move you to tears with their art (Luba Luft), be a reassuring authority figure for years (Garland), or even appear vulnerable and afraid (Pris Stratton), they are actually not people, but rather artifacts designed to fool you about what they are.

It's amazing how many people miss that. The SEO-spammed "grade saver" analysis practically says that the film is right - and from what I've seen, it's probably fair in that the teachers grading you might well think that too. Most misunderstood PKD book by far.

When phrased this way, you're describing real people as well. We all have our various masks, secrets and lies.

> When phrased this way, you're describing real people as well. We all have our various masks, secrets and lies.

You're misunderstanding "Dick went out of the way to rub in that ... robots ... are actually not people, but rather artifacts designed to fool you about what they are." A person fooling others doesn't make that person an artifact.

I understand the idea, but my take is that while PKD was by all accounts a visionary who could see things few others could, I wonder how his opinion on this would have evolved over the last four decades.

Like what makes something an "artifact"? That it is a superficial image produced intentionally by another? Mass propaganda and social pipelining have produced the same phenomenon in humans.

Is it that the robot doesn't "feel" like a human does? Well, now we're in extraordinarily subjective territory, dealing with qualia, and the distinction will become more difficult as neural models continue to evolve and experience the world in the same way we do.

I like PKD's take but I think it's an open question.

Sounds like the book and the film are different works, so people who only watched the film can't misunderstand the book because they didn't read it.

> Most misunderstood PKD book by far.

Most had a different interpretation than you did.

Under the "Fallacy of Intention" theory of literary analysis, even Philip K Dick's intent isn't "the real meaning".

Regardless, "misunderstood" is at best a way to say "unpopular" or "overly simplistic". It isn't wrong, since there are no absolute facts; only opinions.

> This is why warnings of disruption and doom will ring hollow with the public, you’re just promising adventure.

The desires of the public don't enter into it. Like all matters of consequence, the people who control the economy and the government have decided this is happening and didn't take the opinion of the public into consideration.

Well TPTB are actively shaping this opinion, per the article.

You make it sound like a conspiracy but the article clearly argues that this is what the people want. If there really was that much centralized control then crypto would have been much bigger.

Conspiracy? It's just poor pattern recognition at this point to call it a conspiracy. I'm describing the world.

Yes the article argues that. Crypto was not that interesting to people who already have financial control.

> Not only is the populace primed to treat artificial intelligence tools as living peers, they’ve been trained to hate or dismiss anyone who doesn’t

This doesn’t meet my personal experience, but curious to hear if others see something different! The prevailing opinion in the UK at least seems to be that it’s some non-living magical tool assistant.

As this assumption seems to be the basis of the article, I’m not sure on its validity.

I think the article makes a bit of a leap. The "priming" in the passage you're quoting is the depiction of robots in mass-media, cherry-picking the highly anthropomorphized hero-robots (e.g. WALL-E, Lt. Cmdr. Data, the good-guy Terminator in all the sequels).

I think it's an open question if that pre-LLM "priming" will survive actual contact with LLMs or if modern media depictions of robots will change in response.

The priming didn't prepare us to sympathize with robots that were built to plagiarize our art and literature and take our jobs.

Robocop is an interesting example where we sympathize with a cyborg but not the greedy corporate executives. Blade Runner has the same outlook. The Matrix and Ex Machina don't present AI as entirely sympathetic. Etc.

The priming has been a lot more nuanced than simply AI = sympathetic lifeform.

I mean the first point in the article was also:

> “Robots are actually more alive than we are” is a core belief in our society

I absolutely do not think this is a core belief - If anything, the core belief is that 'being a human is special' and that robot's aren't really alive in the same way, because they don't feel human emotion.

I have a feeling we are going to see more media where the opposite arc happens — where the robot is revealed to be actually /not/ as human as someone thinks, and is revealed to be a jerk, actually (like Ash in Alien)

Also adding: the author seems to miss that the stories he references are subversions of the existing trope

Book ad.

No, it’s a blog post on a web page which also has a book ad.

Two book ads to be precise

You can just skip the clearly marked ad bit and read the rest of it. Jason Pargin makes a lot of good content on different channels, and throws in a book plug here and there. Doesn't diminish the quality of his writing in any way.

The rest of it sure does read like "content", yep. There's no substance. It's short. And then there's another book ad.

It reads like someone used an LLM to cough up an article as an excuse to show the book ads. Though I don't actually think it's LLM-generated, mostly just because of the unnecessary ALL-CAPS in the middle, and the overuse of bold, italics, and underlining throughout.

This post makes a critical error. It is talking about sci-fi AI. Artificial Consciousness.

What we have is, while useful for many things (with proper constraints), just prediction engines that behave like intelligence. Generating a sentence when specifically asked to do so is a far cry from independently pondering over your existence then picking up the God of Thunder's hammer.

If anything, the public has been primed to hate AI as we have it now. CEOs are blaming it for mass scale layoffs. Media executives are using it so recklessly they face legal action. Data centers (already an issue for years for environmental impact and noise pollution) are being even more aggressively pushed on communities who don't want them. And the Silicon Valley upper crust are, to sell their AI products, embracing the worst views and uses of them.

AI can be ethical and useful, if we want it to be. But that is not how it is being sold. Increasingly, the public is aligning against it's use. This blog post badly misses that point.

and when AI agents start tasking llms to build better llms on their own that then turn into better agents that then task automated robots to design better hardware then manufacture it and configure it and run agents that start tasking llms to build better llms on their own that etc

at what point does it cease to become a simple mathematical function and start to become an "organism"?

When they start tasking, and don't just "start tasking", you already have a made a leap to something completely different from what we have no. So your question is basically "but what if a frog turns into a prince, at what point is it a prince"?

If that ever starts happening, let us know.

I don't see any evidence that current LLMs even have a path to that.

AI is truly the dumbest technology to emerge in my lifetime.

Hyperbole or just rage bait?

I think it's pretty obvious that genAI is a very smart technology in terms of the implementation, but a dumb technology in that enables devastatingly dumb outcomes. It also enables very smart outcomes, but they are overshadowed by the misery and visibility of the dumb outcomes. Like crypto, but with way more legitimate uses and a wider array of outcomes.

And by "outcome" I mean holistically - not the literal output of a model.

Maybe a smidge dramatic? What I'm seeing is the most money-losing technology being shoe-horned into virtually every system with little to no benefit, but definitely producing a more dystopian future in terms of unemployment and the environment. But even AI fans expect a market correction because they can't hemorrhage cash forever. So in that context, watching employers hire AI-evangelists to force workers to use AI seems even more absurd.

its cope

I've finally started realizing that, while I don't trust AI (or what we're calling AI), what I really don't like or trust are the people behind it. In fact, the whole problem with modern tech is that the people funding it and hyping it are just untrustworthy.

I have less of a problem with Chinese AI than American AI because no one in China is screaming at me to use their AI. No one in China is trying to justify me making sacrifices so their AI can grow. It's like: they don't care if I use their AI or not - it's there, you do you.

Meanwhile, Altman, Bezos, Musk and Zuckerberg (problematic individuals at best) are all over the news proclaiming AI as the Second Coming of Jesus if only I give up water so they can cool their plants, all while yelling "it will all work out if you just trust us!"

Not helping, guys. Not at all.

Also, it's not Intelligent and for the love of God, please stop pretending it is and please stop trying to convince everyone that it is.

Reply on news.ycombinator.com