Well.... touche.
2,382 karma · joined March 18, 2024
GPT-2 was released in 2019, GPT-3 in 2020. I'd say 2020 is significant because that's when people seriously considered the Turing test passed reliably for the first time. But for the sake of this argument, it hardly matters what date years back we choose. There's been enough time since then to see the plateau.
> Most people in society connect AI directly to ChatGPT and hence OpenAI.
I'd double-check that assumption. Many people I've spoken to take a moment to remember that "AI" stands for artificial intelligence. Outside of tongue-in-cheek jokes, OpenAI has about 50% market share in LLMs, but you can't forget that Samsung makes AI washing machines, let alone all the purely fraudulent uses of the "AI" label.
> And there has been a lot of progress in image generation, video generation, ...
These are entirely different architectures from LLM/chat though. But you're right that OpenAI does that, too. When I said that they don't stray much from chat, I was thinking more about AlexNet and the broad applications of ML in general. But you're right, OpenAI also did/does diffusion, GANs, transformer vision.
This doesn't change my views much on chat being "not seeing the forest for the trees" though. In the big picture, I think there aren't many hockey sticks/exponentials left in LLMs to discover. That is not true about other AI/ML.
The wisest move in the chatbot business might be to wait and see if anyone discovers anything profitable before spending more effort and wasting more money on chat R&D, which includes most agentic stuff. Reliable assistants or something along those lines might be the next big breakthrough (if you ask certain futurologists), but the technology we have seems unsuitable for any provable reliability.
ML can be applied in a thousand ways other than LLMs, and many will positively impact our lives and create their own markets. But OpenAI is not in that business. I think the writing is on the wall, and Sama's vocal fry, "AGI is close," and humanity verification crypto coins are smoke and mirrors.
Look, just have courtesy for others and don't argue in bad faith, the snark included. This community came up with the HN guidelines, let's try to follow them more. That's all I wanted to say. All the best.
The parent comment to my original comment mentioned people who are "fake" optimists. I would describe them as jolly or bubbly often, but ultimately pessimistic. That's why their positivity feels disingenuous. It is a facade or a mask. I appreciate their effort, I am simply drawing a distinction between jolly/grouchy and optimistic/pessimistic.
There are tons of comments online that claim emotionally secure and mature people are ignorant or even irresponsible if they're not in perpetual anxiety about the world. The "you should be scared with the way the world is" type. Obviously, what they claim is not true, and it's very unhealthy. But someone who is emotionally secure understands they gain nothing from a neurotic debate online, so they don't call these opinions out. The opinions remain there, unchallenged, seemingly agreeable and prevalent.
Life entrenched in pessimism is also a bit contagious. Though I don't want to be unkind, I have to be honest: there is a limit to how much I can try to help someone with that mindset before they drag me down with them. Something I learned studying field medicine (and this isn't some secret in the first responder world or just general knowledge) is that you protect your safety first. If you arrive in a situation and add yourself to the casualty list, you are just not helping anyone: you get hurt, and you create more work. I think this applies not only to physical hazards, but also to emotional ones. One more depressed person in the world does not make it a better place for anyone.
It makes sense to exercise some caution, and most of our emotional brains understand that. This brings me back to the original point — it's hard for one to succeed anywhere if everyone's avoiding them (best case scenario) or being turned negative themselves (worst case scenario). Much remains to be said about the internet's role in this epidemic of negativity and various depressive disorder crises.
Some things are relatively easy to refund via Apple, but not all of them. It's nearly impossible to get a refund for in-app purchases, gift cards, balance top-ups, auto-renewing subscriptions, redeemed digital goods, and so on. Coming from the EU, if I paid for these things directly with a credit card, I'd be able to get a refund in line with our consumer protection laws (if the item was sold deceptively, if it has a fault, if I am not satisfied for any reason within 14 days) that cover more or less all digital purchases - no problem.
To borrow your words, Apple's system is taking away trust. I like the refunds my European bank offers me because it operates under consumer-friendly laws. I don't trust Apple's refunds.
Anyway, opening any system to more choices for the consumer cannot decrease trust. If the consumer trusts the original payment option, they can use that. If that is not provided, but the customers don't trust other payment methods, the app won't make money. The market will soon negotiate so that the payment methods that customers and sellers find acceptable prevail. Apple fears it won't be their extortion (I mean payment system), and rightfully so. Aside from the Stockholm syndrome, there's very little reason to use it.
Just use a virtual payment processor (PayPal, Amazon Pay, Google Pay, etc) or any credit card directlyh. I mean, you often can't on an iPhone, and that's the whole problem.
If you have a genuine issue with what you bought via Apple's payment gateway and your bank files a credit card chargeback for you, Apple would even indiscriminately ban you. They are hardly the good guy.
They've installed themselves as an arbiter of what can be refunded and what cannot. But by law, the arbiter is the government, not Apple. So there are many, many problems with Apple's approach. The consumer rights are one, but acting above the law itself is a problem.
The Baltics have handled this better with a Smart ID system[0] that also allows cryptographic document signing (like Adobe Sign but non-repudiable by law). It can perform proof-of-human like Altman's orbs, and allows Baltic citizens to file documents with the government, access electronic court records, electronic medical records, e-banking, and similar services.
One downside of the system is that its purpose is to identify a natural person using a service. So whichever service uses it as authentication, it will receive personal data. However, there are third-party sign-on gateways that use Smart ID (and other e-signature methods legal in the Baltics) to authenticate users and only disclose certain bits of personal information to those requesting authentication. In Lithuania, the government operates a service called the E-Government Gateway, and one can easily imagine it could be used for zero-knowledge age verification.
Ultimately, Google seems to be offering a far inferior product at a lot more risk to the user. Once again, their core business is user profiling, associating various user metrics with each profile. A government ID is the holy grail for user profiling. It's sort of like if a wolf offered a sheep-verification service - yes, we could trust the wolf to act professionally towards the sheep, despite it coming out in sheep court several times that it hasn't in the past. But is it wise to suspend disbelief like that? It's better to leave this to independent, expert companies, or even governments.
Looks like face scanning might come to Razer cameras in the future to verify people in multiplayer games and eSports tournaments. Also, for dating on major apps.
What % of the public holds these opinions? Who knows? Maybe someone will conduct a study one day, but 99.9% of internet opinions don't get studied, and 99.9% of people don't post mainly responsibly verified information.
There is so much to say about this, when most of our socialization has moved online. So much to be said.
Apols for the tangent.
I think that could be due to some uncertainty in the market about the future of mobile gaming. Mobile AAA never caught on, and the industry tried it with CoD, Civilization, Mario, Fortnite, Resident Evil, GTA, Assassin's Creed, and others. Only Pokémon Go and Genshin were exceptions on the AAA side due to strong mechanics (Pokémon: ARG appeal, Genshin: gacha profits).
Meanwhile, the small-budget side has become extremely overcrowded by trash-level asset flips, gachas, and ads with vague game elements. Many genuinely talented game developers are asking, "Does my game fit in the mobile markets?" and concluding that the answer is probably closer to no than a yes.
Maybe this is just the natural life cycle of the stores. Moderation doesn't keep up with use, leading to disappointed users and abandonment. Do you remember when we used to check the App Store occasionally for new fun apps during the iPhone 4/4S era? We'd get like an accelerometer app that turns a phone into a silly beer pint, and it'd be quite clever and novel. Or we'd download pedometers, and that was quite a novel and smart tool to track exercise. All because they were promoted on the stores. That's a distant memory for me; I only go to the stores to download the things I already know I need, they have turned into package managers for me.
Now that US courts are doing it more, it seems that corporations abusing their monopoly powers are the problem, not EU laws. But what do I know.
But we know this ideal version of events does not happen often, and large populations of people did in fact become a lot more sedentary as a result of not having to do physical activity in day-to-day life.
Though of course, there is the argument that using calculators didn't make us more mentally inept, so why would AI? But I'm not sure, this seems different.
It's kind of like those romance scams online, where the scammer always love-bombs their victims, and then they spend tens of thousands of dollars on the scammer - it works more than you would expect. Considering that, you don't need much intelligence in an LLM to extract money from users. I worry that emotional manipulation might become a form of enshittification in LLMs eventually, when they run out of steam and need to "growth hack". I mean, many tech companies already have no problem with a bit of emotional blackmail when it comes to money ("Unsubscribing? We will be heartbroken!", "We thought this was meant to be", "your friends will miss you", "we are working so hard to make this product work for you", etc.), or some psychological steering ("we respect your privacy" while showing consent to collect personally identifiable data and broadcast it to 500+ ad companies).
If you're a paying ChatGPT user, try the Monday GPT. It's a bit extreme, but it's an example of how inverting the personality and making ChatGPT mock the user as much as it fawns over them normally would probably make you want to unsubscribe.
It was written with 10% higher expectations, 10% higher bar (it was raised in a transformative way) and 10% fewer sloppy thoughts, by a rockstar 100x commenter who uses enterprise AI to blast their KPIs through the pipeline each day.
Sometimes they even circle back to blast their pipeline several times in the same day, all thanks to AI.
Venture capital accepted by Venmo and Apple Cash. Past performance is not indicative of financial statements. GAAP statements may be disrupted and revolutionized at the sole discretion of the company. T’s and C’s apply.
Have wonderful AI day! :)
After 15 or so iterations, both assistants would keep repeating the same things and find agreement anyway. Sometimes, the chat became unhinged and useless, but 95/100 times, it was agreement.
Happy someone else made it work.