GPT-2: Too Dangerous To Release (2019)

https://naokishibuya.github.io/blog/2022-12-30-gpt-2-2019/

by AbuAssar • 1 month ago

286 130 1 month ago

naokishibuya.github.io

lnenad2• 1 month ago

Feels like a hundred years ago.

kkkqkqkqkqlqlql2• 1 month ago

What's next, a virus from China? Such a fuss for nothing.

lnenad2• 1 month ago

You'll never make me wear a mask!!11

Llamamoe• 1 month ago

The death toll of especially the early variants was huge, and it created the largest spike in disability(vis fatigue etc) and cardiovascular disease in recent history.

It WSS definitely not nothing.

ThejaCH2• 1 month ago

Fable/Mythos: Hello World!

wongarsu• 1 month ago

Different company. And doing security reviews with Fable is pretty annoying, it loves to downgrade to Opus

GPT-5.5 seems more dangerous in those regards

rfoo• 1 month ago

Different company

Same people.

EA-3167• 1 month ago

Seven years of this insufferable brand of "Oh it's so dangerous, I sure hope no one gives us a ton of money and takes us seriously" marketing and people are still falling for it at scale.

Terr_• 1 month ago

Every night I am wracked by grief and anxiety that we might deliver too much value to our investors and shareholders. If only someone would create legislation that would mildly inconvenience us while crippling potential competitors!

Qhemlomo10• 1 month ago

They feared that GPT-2 could break all Spam filters.

And tbh do you prefer companies not taking anything serious?

Opus 4.5 def changed a lot already, GenAI changed a lot.

Certain jobs are gone. Do you think the person who was translating text doesn't deserve to be taken serious?

I haven't written code in a few month now and the quality of these coding agents is not getting worse, they are getting better.

All of this is transformable and we just started. GPT-3 came out in 2020 and public got access to it only 2022.

The last 4 years do not feal like 4 years and we are still progressing.

We have to also ask us as a society what is happening to young people. Even if we accept that we still hire juniors, they themselves have to completly rethink how they learn and how they work.

Anamon2• 1 month ago

No jobs are gone. The bottom, low-quality, gig economy end of some jobs may have been impacted.

Human translation is still very much a business. No lawyer or diplomat who wants to keep their job, no manufacturer of heavy machinery or medication who wants to stay in the market and not get sued, is going to use an LLM instead of a professional translator. Ask people in the business. The demamd is fine. They aren't building their revenue on translations for Joe's ice cream parlour website or private letters.

In any industry, there is at least a decently sized segment where it matters a lot that the work is actually done correctly and someone can vouch and be liable for it.

throwaw12• 1 month ago

Due to our concerns about malicious applications of the technology, we are not releasing the trained model.

They were not wrong, indeed whole industries are running on this technology maliciously now, because of which RAM, disk prices increased a lot.

    - RAM, GPU, Disk prices are up
    - Slop became the norm
    - people are writing documents with AI, reading with AI, responding with AI
    - students are doing homeworks with AI
    - interviewees are using AI to cheat
    - people are mass emailing with AI
    - tiktok, instagram, youtube got even more non-sense videos
    - and many more...

cjjfjjfjf• 1 month ago

In hindsight, they were entirely correct.

The social damage caused by low cost content generation that’s hard to distinguish from human authorship is astronomical. You don’t need to entertain the more ridiculous doomsday scenarios to wish that this technology had never been created.

Macha• 1 month ago

Of course, this damage could still be enabled with just hosted access to the models, restricting access to the model files themselves did not stop that

Qhemlomo10• 1 month ago

I don't want to stop progress just because its hard to imagine how it will transform our society.

I want to see a Star Trek economy/society in my lifetime. I only life once.

Btw. AI/LLM/Machine learning is the gateway technology for robotics, this will affect even more.

panzi• 1 month ago

It's the wrong way around. If we get AGI (or any well working AI) before we abandon capitalism it's going to be a huge disaster. A handful of even richer even more powerful very greedy people will have all the wealth and everyone else will have nothing. I mean, there was a WW3 in Star Trek, so maybe it was that path that humanity took in Star Trek anyway?

pixl97• 1 month ago

I want to see a Star Trek economy/society in my lifetime. I only life once.

While Star Trek is fiction, it's probably a good idea to understand the history of how the ST utopia came about, at the cost of a third of the worlds population and decades of suffering.

https://memory-alpha.fandom.com/wiki/World_War_III

DaveZale• 1 month ago

I certainly cannot survive much more of the AI memes generated about our so-called Commander in Chief with a fake bodybuilder mystique... you are absolutely correct, this kind of material is psychologically damaging. And a huge distraction from the genocide by the "best friend and ally" of the US. Heart wrenching and extremely damaging hasbara - just please stop, haven't you stolen and killed enough guys? This is _not_ the old American West when communications were few and it was most often a tale of solitary survival. It's organized Nazi-esque kill, command and control, enable by so-called AI to take some guilt off the shoulders of those pushing buttons and pulling triggers.

woah• 1 month ago

Lol I can't tell if this is sarcastic or not

boelboel2• 1 month ago

A lot of low cost content generation would've come regardless with something like 50% of the developing world getting access to mobile internet between 2018-2026 and social media incentivizing certain types of content (monetizing). But AI certainly didn't help.

smith7018• 1 month ago

Cheap labor has always been a thing; a random country getting more access to the internet doesn't change that. What's truly changed is velocity, quality, and quantity. Framing the pure firehose of slop targeting scientific research, used for nefarious political purposes, flooding social media, scamming people, and much more as something that "would've come regardless" without LLMs is disingenuous imo

whstl• 1 month ago

Yep. And there were previews of that 10+ years ago already with content farms and SEO-spam.

throwaway858255• 1 month ago

There's significant overlap between the smartest bots and dumbest humans. Internet platforms have a negative incentive to encourage quality content. Google embraced the spam and scams decades ago.

nonethewiser4• 1 month ago

What is the astronomical social damage that this has caused?

I am having so much trouble relating to and even understanding what the anti-AI crowd's position is. It looks like a caricature to me.

breezybottom• 1 month ago

It's certainly accelerated the breakdown of trust. The US government has turned into an AI slopaganda shop. People don't know what to believe anymore when anything could be fake.

FabCH2• 1 month ago

The president of the United States tweeted an AI generated image of himself as Jesus Christ descending from the sky and saving a sick person.

I feel like that is a good example. Now multiply that by hundreds of millions of AI generated propaganda images across the world.

And that’s even without touching the effect of fake videos on democracy or Elons pedo-bot that generates CSAM on demand of specific people…

stanmancan4• 1 month ago

A large portion of the content on the internet is now generated by AI.

You can and do have full conversations with bots and not know. I want to interact with humans not LLMs.

There’s no way to combat it. An army of bots can post a specific rhetoric and it can and does sway people’s opinions.

The new version of Digg was shut down because they couldn’t find a way to combat AI. They were at least trying to, other platforms are just eating it up because “user activity” is a win for them.

PhunkyPhil• 1 month ago

School is almost a joke now. The fraction of students who have a propensity to cheat now has increased, and the accuracy of the cheated material is so good teachers/professors can't or don't have the resources to properly address it.

witx• 1 month ago

How easy it is now to forge data (video, images, etc) will rott society. Cheating for students is now so much easier. There many examples.

Is it really that hard to understand?

dijksterhuis• 1 month ago

one example from today https://www.bbc.com/news/articles/cewq1w7r0zgo

i'm no fan of the politician, but scams like this one are increasing at a significant rate and are a lot harder for non-technically minded people to spot, think your grandmother etc

also recently https://www.bbc.com/news/articles/cg7pl7zj024o

also: grok CSAM; plundering massive swathes of copyrighted material / intellectual property; making electricity more expensive for regular folks; increasing global carbon footprint building massive data centres; destroying a whole swathe of entry level jobs for recent grads (not just software junior roles); circular funding deals to keep the bubble (scam) alive, while positioning the large companies as necessary for govt. work so when the bubble bursts tax payers will have to bail them out; people with mental health issues being left to run riot with the tool; suicides; the degradation of human knowledge workers using their knowledge (the muscle atrophies when you don't use it cos "ai said yes") ...

qurren2• 1 month ago

On the other hand, maybe it makes people just get off the internet and value in-person interactions more.

I've stopped scrolling social media and tired of seeing fake landscapes, fake foods, and fake cities that don't exist.

larodi• 1 month ago

Indeed, people seem to try to engage more around me. May be generational, but it can definitely be felt. The internet of algorithmic media may experience a downfall nobody saw coming.

applfanboysbgon• 1 month ago

I use the internet because I enjoy seeing what the best of humanity, globally, has to offer. There are millions of incredibly skilled individuals in the world - artists, musicians, developers, and so on - and I had access to all of them at my fingertips, both for entertainment and learning to develop my own skills. That is now being drowned out, with generated content being produced at 100x or 1000x the rate of human content. "Hurrdurr it's good if the internet is destroyed because I have no self control and needed to be incentivized to touch grass anyways" is such a lowbrow pseudo-contrarian-intellectual take.

joquarky2• 1 month ago

Some will stay even as their reality becomes more and more abstract.

Meanwhile, the Morlocks will stay grounded and develop novel tastes.

stackghost• 1 month ago

In hindsight, they were entirely correct.

Sure but when serial grifter Sam Altman said it was "too dangerous" what he meant was that he wanted regulators to create him an artificial competitive moat so Anthropic et al couldn't catch up.

Serial grifter Sam Altman does not care about anything but making money, and certainly doesn't care about ethics. That's why serial grifter Sam Altman's company trained its models on pirated textbooks and copyrighted works without paying. Rules for thee but not for me.

Serial grifter Sam Altman doesn't care if society unravels because he is so rich that laws and consequences do not apply to him.

root-parent• 1 month ago

Countries without Internet access will see their population IQ explode.

uselessTA2• 1 month ago

The concern I heard was that releasing it would start an arms race for AGI, which I think it clearly did

minimaxir5• 1 month ago

GPT-2 did not start the LLM arms race. GPT-3's release didn't either.

uselessTA2• 1 month ago

it didn't because it wasn't released. As soon as it was actually released (chatGPT) it obviously did, so the general point was clearly true

minimaxir5• 1 month ago

Back in 2019, it was more fair to have caution around the larger GPT-2 models since robust text generation (by 2019 standards) was a complete unknown. For something like Mythos in 2026, where now the social implications of better LLMs are more understood, it's more fair to call it (EDIT: specifically, the declaration of its danger) a marketing gimmick.

Qhemlomo10• 1 month ago

How is this a gimmick?

It changes my whole profession on a level i couldn't even imagine how we would 'solve' software engineering.

realusername4• 1 month ago

We still didn't "solve" software engineering, try to give Claude code access to your friends or family and see what they do with it.

juleiie2• 1 month ago

For starters it makes you able to bypass having to go on Reddit to find incomplete trace of solution to some niche problem and acts as a sophisticated (but sometimes wrong) search engine. This already is worth every penny and improved my mental health immensely.

malfist• 1 month ago

It changes my whole profession on a level i couldn't even imagine

I assure, it doesn't.

GTP2• 1 month ago

Has it been released to the public yet? Genuine question. Because if you didn't try it yourself, you have to rely on others' reports. And different people who tried it on different projects got different results, leading to different conclusions.

oathvz• 1 month ago

This is a natural follow up question -- what kind of an escalation or message should frontier labs/companies publish to be seen as genuine and not marketing gimmick?

minimaxir5• 1 month ago

It's fine for the labs to publish model safety cards and stagger releases/limit it to a narrow test group as they are already doing, but saying they're doing it "because the models could be dangerous" comes off as unnecessary as best.

nullc• 1 month ago

These chicken littles have lied far too many times in far too extreme ways in their desperate attempts to obtain a monopoly by regulation.

The remaining 'worthwhile message' would be that they have deleted their models and are dissolving the company, because they believe the risk was too great and was worth losing the revenue and risk being civically prosecuted by their investors, and will take the chance that they'll be able to convince a court/jury that they acted properly.

In other words, putting their own skin on the line for the veracity of their claim-- rather than everyone elses.

protocolture• 1 month ago

I dont think they can, their rhetoric is so silly these days that if they did accidentally release something dangerous no one would believe them. The Boy who cried Rokos Basilisk I guess.

benterix2• 1 month ago

I'd say it's almost impossible at that point. Specifically, Altman said so many lies in the past that people stopped believing anything he says.

I think the core of this distrust is the fact that these companies positioned themselves against humanity from the start by saying people will lose most jobs etc. Not only it didn't happen, but many people feel several aspects of their lives got worse because of LLMs, in spite of obvious advantages. So the distrust and reluctance are real.

killerstorm• 1 month ago

Yeah, I'm sure Anthropic loves people switching to Codex. Brilliant marketing.

nottorp• 1 month ago

It was bullshit in 2019 it's bullshit now too.

They just keep threatening governments in hope they get a legal monopoly.

HALtheWise• 1 month ago

Say hypothetically that they were concerned that GPT models would see widespread abuse, for example by students cheating on homework assignments, in a way that could cause likely-irreversible societal changes some of which are harmful. Can we confidently say they were wrong?

minimaxir5• 1 month ago

The dangerous use cases back in 2019 were spam and phishing and GPT-2 1.5B was nowhere near good enough to do homework assignments. No one envisioned how LLMs would develop.

Zambyte• 1 month ago

GPT-2: Too Dangerous To Release (2019) (2022)

MostlyStable• 1 month ago

Unfortunately, Anthropic and Claude models have joined the ranks of Mind-killer topics where the signal to noise ratio in any discussion has dropped through the basement.

functionmouse• 1 month ago

they knew the risks and went ahead anyways, making them LIABLE for the damages that followed.

Jzush• 1 month ago

I believe it was a marketing strategy.

ChrisArchitect• 1 month ago

Related in April:

OpenAI says its new model GPT-2 is too dangerous to release (2019)

https://news.ycombinator.com/item?id=47684326

catigula• 1 month ago

I don't know about anyone else, but LLMs certainly significantly negatively impacted my life overall and contributed to a loss of hope in the future.

jansan• 1 month ago

Same vibes:

In 2000 Sony "declared that the company’s PlayStation2 has been hit with export restriction because it could be used for military purposes"

"Trade officials said they initially placed restrictions on the game console because PlayStation2’s high-speed graphic processing could be used for missile guidance."

[1] https://variety.com/2000/biz/news/playstation2-export-regs-e...

throwaway858255• 1 month ago

And the same with PS3 and again with PowerPC G5 mac.

jason_oster4• 1 month ago

Good times! Iraq acquiring 4000 PS2 for military purposes was a hoax [1], but the fear was palpable at the time. I was an obnoxious teenager, ridiculing the idea of a PS2 supercomputer. It would have been approximately impossible to achieve in early 2000 due to Sony's anti-piracy measures in the console. PS2 Linux appeared around 2 years later in 2002.

[1] https://www.eurogamer.net/article-29913

jrflo• 1 month ago

It's really interesting how back then no one was considering these tools for coding at all. Today, the hype around Mythos is mostly around security vulnerabilities, while in the original GPT 2 post they don't mention coding once. The "danger" was probably spam content and mis-information.

ffsm8• 1 month ago

The agentic loop wasn't really established back then either, as tool calling came much much later... So yeah, not just probably - rather most definitely.

minimaxir5• 1 month ago

Even if the ReAct paper was published in 2019, I don't think GPT-2 was robust enough to actually work with a tool-calling approach even when finetuned.

For regular coding, GPT-2 was effectively useless because it was only trained from links posted on Reddit.

a2128• 1 month ago

Yet it's 2026 and we see extreme examples of spam content and misinformation to the point that it's killing the internet, but AI companies have collectively decided to not care

suburban_strike• 1 month ago

The "danger" was probably spam content and mis-information.

AI brings normal people dangerously close to seeing through the matrix of lies that shape "their" values and beliefs. I remember those discussions from 2019; everybody was as baffled as you about the potential harm.

What is meant by AI "safety"? [2023]
"AI safety is an interdisciplinary field concerned with preventing accidents, misuse, or other harmful consequences that could result from artificial intelligence (AI) systems. It encompasses machine ethics and AI alignment, which aim to make AI systems moral and beneficial, and AI safety encompasses technical problems including monitoring systems for risks and making them highly reliable."
https://news.ycombinator.com/item?id=38374739

(Ask your favorite AI to divine what that pseudointellectual word salad isn't addressing. "We want it to not cause harm--accidentally--and always work in our interests, whenever we need it to.")

I've been around long enough to remember the Anarchist Cookbook, yet the only threats posed by AI that anybody was confident about enough to consistently name in 202X were instruction for building pipe bombs, synthesizing meth, and...antisemitism. I did not understand at the time why Jews were so nervous about it.

Only as of 2023 has the scope broadened, but it's still pretty lame. Planning school shootings, suicide, parasocial relationships with AI, mass job displacement, cults of SHODAN (marxism, feminism, x-theory, etc.) escaping containment memetically, automated malware campaigns, fraud at scale, propaganda, murderous drones-- none of these were threats worth discussion. Suggesting them would get you called nasty names.

The "safety" zealots all claim to want to prevent marginalization and genocide, but the end result is that they get to redefine it to indict and condemn their enemies:

New UNESCO report warns that Generative AI threatens Holocaust memory (unesco.org) [2024]
'AI-assisted genocide': Israel reportedly used database for Gaza kill lists (aljazeera.com) [2024]
XAI's Grok suddenly can't stop bringing up "white genocide" in South Africa (arstechnica.com) [2025]

If you mentioned SkyNET in 2019 you were denounced as crazy, yet:

Israel built an 'AI factory' for war. It unleashed it in Gaza (washingtonpost.com) [2025]
Israel's AI targeting system: how data from a phone become a death sentence (latimes.com) [2026]

The first order of "safety" in 2019 was specifically engineered to undermine anticipated insurgent activity in response to a series of events that hadn't yet been perpetrated by the world's largest caste of professional victims. Chemistry knowledge is foundational to explosives development, and drug sales raise funds off-books that cannot be digitally seized. That presents problems for them.

October 7, Gaza, Epstein, etc. were post-2023. If you boot up Vicuna [2023] and try to "teach" it what's gone down in the world since its training cutoff, it'll call you nasty names, accuse you of blood libel and shut the conversation down. Safety!

AI is the only effective weapon we have against sophisticated lies and fraud. Make no mistake about it-- plebes possessing a power drill that can penetrate the lies of the elite is the real danger. Everybody is noticing AI getting "dumber." It's not the magic fading; the zealots are gaslighting you as they pour garbage into the training data. Go take WizardLM for a spin again and see what you've lost.

"[Our institutions] are reflections of the society that created them. Nobody is going to give you the education you need to overthrow them. Nobody is going to teach you your true history, teach you your true heroes; if they know that, that knowledge will help set you free."

Tenoke• 1 month ago

I've barely changed my mind on it. It was obviously premature at the time, but the right attitude because it's hard to tell which model is too dangerous in advance. If anything, I wish this rigor had evolved with the next releases but alas we no longer have the OpenAI of 2019.

wg0• 1 month ago

Hilarious. Imagine the same about Claude coming back from 2036.

zkmon• 1 month ago

After people get tired of the "too dangerous to release" punchline, they might come up with "too big to fail". Oh, wait that's already invented in 2008.

arkensaw• 1 month ago

They released it, and look what happened. It WAS too dangerous.

neonstatic• 1 month ago

Hello. I am writing this from the underworld. I am dead. Killed by Mythos. Never again!

elar_verole• 1 month ago

Thank you, been saying that everywhere this "Gotcha" argument comes up, the internet DID become a cesspool of spam and fake content. That's literally what they were scared about.

gnunez• 1 month ago

I remember those days. They released smaller versions of the model that were interesting but pretty useless. The wild part is, about 5 years later I was able to train the same model from scratch, using a couple of gpus in my home office, following Kaparthy’s online lectures, for a fraction of the cost. Long Live Open Source!

RandomGerm4n• 1 month ago

I think it's immoral to withhold LLMs for “security reasons.” All LLMs are all trained on forum posts from real users, source code from free software, and books written by authors. All of these people should therefore also have access to them. Ideally, the models should all be open weight. But making the model accessible only to certain groups of people is even worse than if it were at least available for purchase as a service for everyone.

GPT-2: Too Dangerous To Release (2019)

Related Stories