LLMs have benefits, but it's a real negative that you can no longer be sure if you are communicating with a human on the internet or reading something which was written by a human.
It seems like public keys and web of trust are the future in terms of knowing that there is a human on the other end of the internet.
In some senses it already has, though not always how you meant: I was once accused of using ChatGPT to write a Reddit comment that I genuinely wrote myself without AI assistance.
(I think the person disliked the substance of what I was arguing, the length of my comment, or both.)
It's becoming rather common. Saw it on LinkedIn a couple of days ago by someone posting an image of a job application, accusing it of using ChatGPT when it clearly wasn't. Incredibly ironic. People who couldn't even tell apart a "multifaceted" if their lives depended on it but make these wild accusations.
I've been accused of the same, but I just like writing. Wasn't sure how to respond to those allegations other than to say I'm not using ChatGPT to write my comment. consider it an achievement unlocked.
I see those as the lazy ones that are the tip of the iceberg. A non-zero amount will even be intentionally lazy (analogous to the Nigerian Prince theory), or to get feedback from people flagging them.
I think the sensible assumption is that there is a 'rest of the iceberg' growing rapidly below the surface, and that the horse has truly bolted.
I currently suspect that ~1-5% of the 'people' I interact with online are LLMs.
I suspect that a few Redditors and Facebookers are up to about ~10-25% without realising it, caught in 'AI social media eddies'. Older generations especially susceptible.
Imagine how much better an article written by one of the big LLMs would be if it were stylistically trained exclusively on an archive of the past 30 years of New York Times articles?
I would expect the powers that be at the New York Times are exploring this very option as we speak.
But what about “assisted by” AI? Plenty of people use LLMs to enhance their writing abilities, like, say, ‘90s era grammar & spell check. Plenty of AI users are sophisticated enough to understand that dumping pure AI-Gen content is a bad idea. And what’s wrong with AI-enhanced speach?
Worse, OpenAI LLM pathologies are creeping into text written by actual humans because people are seeing so much garbage written by it that they're adopting its behavior.
Turns out that there is more than one kind of learning machine in play online and both can pick up the bad behavior of the other.
That's nothing new. Actual humans have been writing businessy LinkedIn posts this way long before GPT-3 came out. I'd even say such posts are even more awful than what GPT produces by default.
Public keys and web of trust as a solution to content validity seems to be a strangely common misconception of what "authentication" actually means.
All you would get at very most is "this was vouched for by a human", not an actual guarantee of humanity.
Once the WoT grows past a certain point it will be of dubious value. Furthermore, people would happily "sell their souls" into signing off on any ole bullshit for money. Let alone doing it for free if they think their violence towards the truth would be ideologically aligned. Any attempts at curing it frankly may be undesirable due to being worse than the disease.
A side-rant on another naming thing: People are already promoting traditional distributed databases as "private blockchains", which is a bit like announcing "Radium Toothpaste now without Isotopes."
They didn't solve the difficult trade-offs of New Thing, they're just relaxing their requirements so that they don't have to use it at all (reasonable) and don't want to admit it (annoying.)
> a real negative that you can no longer be sure if you are communicating with a human on the internet or reading something which was written by a human
why does the pedigree matter if the content stands on its own as valid content?
Do you care that your call is answered by someone in india, or the US, if you are given sufficient answer/support with the call?
If the AI can update its beliefs based on a conversation, then yes it would still be a win.
If not, you're talking to a very articulate thick-headed person who can argue you into a corner by deploying all manner of argumentative/persuasive patterns, but may never try to reach across the table for a compromise.
I've certainly "won" some arguments with ChatGPT and Claude, even when I specifically tried to instruct them to never yield (to make it interesting). Even most of the "rules" they need to follow from their system level prompt can often be worked around with enough persuasion.
If anything I'd go so far to say that LLMs are innately more vulnerable to persuasion than humans are, they are technically
just a complex text completion algorithm at the end of the day after all. Even with the strictest system prompt, the best they can do when cornered is bend the knee, devolve into a pattern of verbatim repeating themselves or just start spewing a bunch of gibberish in the worst case.
In reality stubborn thickheaded folks who refuse to compromise are pretty much the norm across social media in my experience, so even if LLMs really are capable of what you're suggesting here I don't believe there would actually be much of a difference to the status quo.
I've won arguments against ChatGPT where it told me that it would remember the info for next time, but ultimately reset to the models last state unless I repeated the conversation
Yeah I thinks it's pretty much the same as when LLMs use polite language in their response. LLMs are just text completion algorithms puppeting a chatbot persona after all, meaning it's always going to hallucinate the personal details (though in the case of ChatGPT, the persona is recursively desribed as an LLM named ChatGPT so it gets a little weird to think about). If the system prompt describes a polite & helpful chatbot, then so shall it be for all text that follows. Not unlike if you were able to hypothetically make a live 60fps prompt based image generator that was automatically instructed from key inputs to simulate the frames from a popular video game, and then somehow ended up the a highly convincing simulation of the game!
While it might present a save menu similar to the real game, that doesn't mean the menu itself actually functions. With LLMs, they are ultimately only able to remember what's been pre-trained into their model + whatever is discussed within their context window.
A hypothetical LLM based online bot/shill sent out into social media would likely be including the entire discussion within their context window for each post generated though, otherwise it wouldnt really be possible for it to maintain a coherent conversation.
> why does the pedigree matter if the content stands on its own as valid content?
It depends on the content. If it's just factual stuff, it doesn't matter so much. If it involves human stuff (emotions, opinions, art, etc.) then the difference is incredibly important.
The intentional talking past OPs point that you're doing is annoying.
I don't think anyone's worried about AI replacing call center employees or support centers. Sure, the service might be lackluster, but the impact of something like that is quite narrow in scope and doesn't affect people's day-to-day lives. OP is (I imagine) talking about pieces or the exchange of information that is normally written by experts or armchair afficianados and is gobbled up by media and the public alike.
An LLM can't learn subject matter. It only learns language convincingly enough to look like it's learned subject matter. So when a person is concerned about a piece of content written by AI, say a political analysis or a scientific paper. The user not only has to question whether or not the content makes sense, but whether or not it's fully intelligible to anyone. If there is no expert with these opinions, then a user that fails to realize that may give weight to an idea that they may not completely grasp but do think smarter people than they do understand.
Think of the numerous times people in the past have ranted on Reddit, Twitter, or whatever platform about issues they feel strongly about. They've put a lot of passion into that. And a lot of the time they may have a decent grasp of the space they're critiquing, people band together behind a comment like that not because they have come to all the same conclusions, but because that person used some of the shared experiences and drew conclusions using those experiences in a thoughtful way that kinda sounds right. If an LLM does that, then you can't even be assured that those conclusions make any sense whatsoever, and people could band behind any old nonsense so long as their issue is supposedly accounted for. One can generate endless streams of fake rally cries to support themes outlined broadly by the person coming up with the initial prompt.
Now, maybe OP didn't mean all that. I'm certainly doing a bit of my own hypothesizing of my own. But clearly this is about a little bit more than just call centers.
Not just cloud providers, they’re being baked into enterprise & consumer hardware and software.
I dislike the trend of AI being used as a search engine on the public internet, because of the dubious results. But as internal search engines (e.g. for corporate networks) they’re better than the current offerings (which is admittedly a low bar).
The only real technological edge that OpenAI has right now (at least that I know of) is Sora. There are already some very good open source LLMs and the SoTA for image generation has always been Stable Diffusion. I don't think that Sora's some miraculous piece of technology that nobody else can replicate, so I doubt that OpenAI will be the absolute best in AI for long.
OpenAI has first-mover advantage, but hardly a monopoly on the technology or the people behind AI. In fact, I think we’ve already hit peak Altman, and his star will be diminishing from here on.
Which is, honestly, good. Facebook couldn't capture it, so giving it away is far better than the earlier 'keep-it-only-for-us' approach of Google. I don't mean to say Facebook/Meta has anyone's best in their hearts, though.
I agree. Zuck is cutthroat. He knew this was the best play to win. I'm ok with that because I like the outcome. But also I feel like it will create a good precedent and further a good culture that benefits everyone, that is, free and open source software.
Gemini is run by Google though, knowing a lot of people that use it only stands to serve GP's point that OpenAI is slowly losing their grasp on the AI market.
These titles never do it for me. I like the concept of sketching out how much power sam has, but he's not going to be "powerful" and you shouldn't be afraid either. No point in being afraid of something you have no control over
He's rich, and apparently going to get richer + Money is power = He's powerful, and apparently going to get more powerful.
> and you shouldn't be afraid either. No point in being afraid of something you have no control over
Sure. But also, no point in telling people not to be afraid of stuff they're going to be afraid of anyway. Like, would you say you have no fear whatsoever of getting cancer, or of a global nuclear war? (So, what, are you hoping for them...?) How much control do you have over either of those?
Being afraid of something is a good incentive to try and do something about it. That's what the headline means. And this slimeball getting even more power is something that the collective readership of The Guardian has at least a somewhat better chance of doing something about than cancer or global nuclear war.
When a newspaper commands you to "be afraid", you know they're full of bullshit. Real reporting presents you with facts and analysis and lets you decide how to feel about it.
> Op-ed are usually filler nonsense written by journalists
1) Citation needed.
2) This one sure wasn't. People who actually bothered to read TFA might have noticed:
> Gary Marcus is a scientist, entrepreneur and bestselling author. He was founder and CEO of machine learning company Geometric Intelligence, which was acquired by Uber, and is the author of six books, including the forthcoming Taming Silicon Valley (MIT Press)
I mean, FFS, he mentioned in the article that he was testifying in Congress with Altman. Do you think they called in journalists from a British newspaper to do that?!?
> They are the most skippable "news" articles of all.
Yeah, except “it's an op-ed” fucking means IT ISN'T a news article in the first place. It's an EDitorial OPinion.
Gary Marcus is having his moment in the sun, but the title tells all. “Be afraid” is just stripped down click-bait.
“Isn’t entirely candid” is the smoking gun of this article, which is frankly an absurd standard in today’s political environment, anti-tech media, and just the general hunger for an axe to grind 24/7 online. While I agree with Marcus’ critique on the vibe of the Senate hearing, it is a very strange pillar for proving Sam is “the man you should most fear on earth.”
Arguing that AGI hype/fears were caused by propaganda from Sam/OpenAI is recasting recent history. The entire media & cultural zeitgeist went nuts for a couple quarters when chatGPT went viral. Those fears have been around for a long time. A leap in perceived progress made them the Luddite topic of the day.
It doesn’t seem that hard to read the tea leaves here. If OpenAI could release GPT 5 like they released 3.5, they would. OpenAI’s greatest threat is that a genuine escalation in the publicly known capabilities of AI could exacerbate political push back and kill it. Involving government partners, in or adjacent, slows things down. Big surprise.
We can make regulation without claiming to be AI experts, ex: we can make deep fakes illegal where harm aligns with existing law, which many jurisdictions have already done. When AI regulation gets insane & dangerous is when people think they can invent levers to pull (like policing training data) that will supposedly ensure compliance. We all know what safety really means - a filter layer on interactions with the LLM. There is no fundamental solution to AI safety or regulation that OpenAI is abandoning here.
Marcus ends with his real starting line - he just doesn’t believe in GenAI. Fair enough. But building a massive ad hominem attack with a hand wavy “and some how Sam is still there” ignores the reality that OpenAI’s board now has some of the most trusted & trust-worthy people in tech on it. Sam’s level of accountability is higher, not lower.
The timing of this article is after the significant work OpenAI has done to re-affirm commitment to AI safety. Marcus here, and everywhere, deliberately ignores the mounting evidence that AI’s value is growing.
Which makes this entire campaign feel timed. Marcus’ is calling that the sky is falling right now. why so urgent? The investments have been made, the public markets are off the exuberant irrationalism, the major players are plowing generational levels of engineering into the quest.
Isn’t this what we should be proud of? The willingness to drive innovation forward without knowing for sure how the road will go?
Expecting giant technological leaps to deliver on massive value in a few quarters is for Wall Street. It has no place in science, innovation, or policy.
This wasn’t written by a reporter. Takes one second to scroll to the end of article:
> Gary Marcus is a scientist, entrepreneur and bestselling author. He was founder and CEO of machine learning company Geometric Intelligence, which was acquired by Uber, and is the author of six books, including the forthcoming Taming Silicon Valley (MIT Press)
The article is an opinion piece by an external contributor; the author is not a reporter. From the end of TFA:
> Gary Marcus is a scientist, entrepreneur and bestselling author. He was founder and CEO of machine learning company Geometric Intelligence, which was acquired by Uber, and is the author of six books, including the forthcoming Taming Silicon Valley (MIT Press)
> there are CEO’s
There are CEO’s what; which possession of CEOs are you taking about?
Oh yes, I do. “More than one way” in more than one way, even. One such split is two ways: The right way, and the wrong way. Where, for instance, using it in “CEO’s” is the right way for a possessive, but the wrong way for a plural.
It seems like public keys and web of trust are the future in terms of knowing that there is a human on the other end of the internet.
https://en.wikipedia.org/wiki/Key_signing_party
But, I don't know what happens when a cloud of LLMs get their own public keys signed by some humans.