7 ms·
Do they not have anyone sensible left in the room? Like, you’d expect that at some point someone would have said ‘these are comically terrible, we cannot allow
by rsynnott 2y ago
Do they not have anyone sensible left in the room? Like, you’d expect that at some point someone would have said ‘these are comically terrible, we cannot allow them to see the light of day’.
- motohagiography 2y agoIt's like people have already forgotten why people used the non-human beings in Westworld. One of the first things we're going to use them for is amusement and having them represent anything sentimental or sensitive is going to make them a target like this. It was irresponsible. There's something essentially wrong with Meta and Google where they can do tech but not products anymore. I'd argue it's because the honest human desire that drives a product dies or is refined out of initiatives by the time it gets through the optimization process of their org structure. Nothing is going to survive there unless it's an arbitrage or using leverage on the user, and the things that survive are uncanny and weird because they are no longer expressions of something someone wants. These avatars ticked all the boxes, but when they arrived, people laughed at them because objectively they were bureaucratic abominations.
- ToucanLoucan 2y ago> There's something essentially wrong with Meta and Google where they can do tech but not products anymore. Because AI is the hype thing and adding AI to your thing makes the stock price go up, because investors and stock evaluators don't know what AI is, nor do they care. They just know it's the hot thing and what you put in your product to make it look better to Wall St. Meta AI, Google AI summaries, Apple Intelligence are all hilariously half-baked features designed to connect users with ChatGPT so the line goes brrrrrrr. Them being helpful, useful products that solve problems for users is a distant, distant next place to that.
- bloomingkales 2y agoI think Google and others are too distracted in collecting enterprise coin at the moment. They have a perfectly good consumer product in NotebookLM, but at the moment it has the quality of something an intern made.
- twoodfin 2y agoWhy would consumers pay for NotebookLM? It’s basically an (impressive) party trick.
- Spooky23 2y agoWhy would you say that? I used it as a study guide. Super useful. Stuff like uploading 8 hours of audio and asking it: “Generate an outline of topics that the instructor said were important to remember.
- scarface_74 2y agoNotebookLM is more than just podcast generation. I came into the middle of consulting project where there were already dozens of artifacts - SOWs, transcripts from discovery sessions, client provided documentation etc. I loaded them all up into NotebookLM. I started asking it questions after uploading it all to NotebookLM like I would if I were starting from scratch with a customer. It answered the questions with citations. And to hopefully deflect the usual objections - we already use GSuite as a corporate standard, NotebookLM is specifically allowed and it doesn’t train on your data.
- oceansweep 2y agoAs someone who's built something like it in their free time as a hobby project ( https://github.com/rmusser01/tldw https://github.com/rmusser01/tldw), could I ask what would make it a professional product vs something an intern came up with? Looking for insights I could possibly apply/learn from to implement in my own project. One of my goals with my project I ended up taking on was to match/exceed NotebookLMs feature set, to ensure that an open source version would be available to people for free, with ownership of their data.
- BoorishBears 2y agoI'm going to challenge you to put that first screenshot into ChatGPT/Claude and ask them why it looks like something an intern came up with vs a professional product. I'm not saying that as a slight or an insult, but right now the screenshot looks like a Gradio space someone would use to prove out the underlying tech of a professional product, not a professional product (unless you literally mean professionals are your target users as opposed to consumers). I think an LLM would be able to very quickly tell you what most product builders would tell you at this stage. - Also one of the key enablers of NotebookLM is SoundStorm style parallel generation. Afaik no open source project has reached that milestone, have they?
- sonofhans 2y ago> There's something essentially wrong with Meta and Google where they can do tech but not products anymore. I'd argue it's because the honest human desire that drives a product dies or is refined out of initiatives by the time it gets through the optimization process of their org structure. This is a really nice insight. I hadn’t put it together this way before. It’s similar to design-by-committee, or movie script-by-committee (like Disney-era Star Wars movies, or a Netflix focus-group-driven script). The layers of bureaucratic filter have rubbed off all human influence, and all that’s left is naked profit motive. The worst thing IMO would be if LLMs became able to convincingly fake these emotions. They’d become emotional pillows for the inhuman manipulative ambitions of their parent orgs.
- baxtr 2y agoI wonder why so few companies employ Pixar’s Brain Trust method. You get into a room with a bunch of key execs who are not part of your project and then they tear your product/movie/idea apart. You don’t have to implement any of their suggestions but since these are seasoned execs they see flaws readily. The key is that the process encourages candid feedback without hierarchy or defensiveness. This goes on for a couple of iterations until the idea is ready for production. I think Apple has something similar.
- thereisnospork 2y agoImo it's due to people have a hard time separating the worth of their idea with their self-worth. So criticism of their 'baby' breeds negative sentiment which creates a challenging political environment for criticism, which would need to be ameliorated by strong organizational culture.
- Capricorn2481 2y agoI think the commenter is saying, with a brain trust, you don't share organizational structure or hierarchy. Neither party has enough interaction for a political environment to form. You can be offended, but you won't see them again and you're not forced to implement their ideas
- PKop 2y ago> It's like people have already forgotten why people used the non-human beings in Westworld This is fiction though. Perhaps we didn't forget about that fake thing and we're critiquing this real thing that exists. How can you take a demonstrative position on "what we're going to use them for" and defend it with someone's contrived story? > because objectively they were bureaucratic abominations Yes, this is why criticism of this real thing isn't countered by claiming people "forgot" what occurred in a fantasy script about things that don't exist.
- motohagiography 2y agothis misunderstands what fiction is. for people with the capacity, it's a tool for reasoning about hypotheticals and counterfactuals. sometimes its fun, but mostly it's serious. for the people who get it fiction is a public discourse about possibilities. to me seeing it as arbitrary would be like watching golf and thinking it's random. there's a literal mindedness or incapacity for abstraction I can't apprehend in that.
- frereubu 2y agoThe point is that a fictional story is just one of a host of possibilities, so you can't base decisions off what happens in it rather than the other n-1 possibilities.
- jdbernard 2y agoThe problem with using fiction as a discourse about possibilities is that fiction is governed by the rules of the author's mind, not the rules of reality. So the fidelity of the model being used to drive the discourse is directly dependent on the congruence between the author's internal model of the world and reality, which can often be deceptively far apart. This is especially bad when the subject is entertaining, because most of us read fiction for entertainment, not logical discourse. So we create scenarios that are entertaining rather than realistic. And the better the author is the more subtle the differences are, but that doesn't mean they go away. It feels like a somewhat common experience in my life that I'm discussing some topic with somebody, and I have subject matter expertise based on actual lived experience, and as the conversation goes on I discover that all of my conversation partner's thinking about the subject was done in the context of a fictional world which misses key elements of the real world that lead to very different conclusions and outcomes. This is not to totally discard fiction as a way of reasoning. With regard to hypotheticals beyond our current reach it is often the only way to reason. So it's valuable, but we have to keep in mind that a story is just a story. Hard experience trumps fictional logic any day. And I can't assume that the same events in real life will lead to the same outcomes from a story.
- loeg 2y agoFictional dystopias aren't the real world. Westworld is just television.
- johnvanommen 2y ago> There's something essentially wrong with Meta and Google where they can do tech but not products anymore. I'm listening to "Masters of Doom" on Audible. It's about the creation of Wolfenstein 3D, Doom, etc. Great book. Something that's interesting about it, is that Id Software seemed to largely reject the idea of having a game where the players 'connected' to the characters or the story. They were laser focused on: * Carmack creating a game engine that was the best in the world * Romero and crew making the gameplay as fun as possible But it sounded like they had folks on the team who wanted to make a story where the players could 'connect' with the protagonist. He was fired. Some dude named "Tom." Facebook, weirdly enough, seems to have the issue. Which is particularly odd considering that's their product! I haven't played games in ages, but when I saw "Half Life" for the first time, it felt nearly as "revolutionary" to me as Wolfenstein 3D was.
- NegatioN 2y agoI feel like the way Romero and Carmack wanted to make the player "connect" with the game was different than what Tom had in mind (from reading the book). Tom wanted elaborate lore and story-telling, while the rest wanted to make the game experience what the player connected to. The instant reaction to your input, seeing your bad-ass character (and by extension yourself) inflict awesome damage on the world. This to me is more of a conflict of _what_ the player should connect to, as opposed to not wanting the player to connect at all.
- namaria 2y agoThat's the price for Mr Zuckerberg's ownership structure. On the one hand, no one can dispute him. On the other hand, no one can dispute him.
- laweijfmvo 2y agothis is exactly what happened
- roominator 2y agoI could also see this being some executive trying to justify the word "AI" in their title with an initiative that should "make number go up" wrt to engagement or something. "When we have more users engagement goes up, let's just _make more users_".
- MrMcCall 2y agoAnd he is personally responsible for so much misery on Earth that he can likely never do enough good to make up for it, not that he's going to try. The reason he never looks happy is because he is never happy. Same with his fellow oligarchs; they can get pleasure in bunches but never happiness, they can be smug but never have peace. Happiness is what happens as a result of your helping others become happier; there is no shortcut, there is no other way. Zuck has only ever worked for himself and his "peers". History is riddled with those losers, the richer the worser. It is not just human nature but the nature of the universe vis a vis our human responsibility as choosers of goodness or its opposites.
- drdeca 2y ago> Happiness is what happens as a result of your helping others become happier; there is no shortcut, there is no other way. Zack has a comic on a similar concept: https://www.smbc-comics.com/comic/2009-11-18 https://www.smbc-comics.com/comic/2009-11-18 Of course, the exact line of reasoning doesn’t quite work for the different formulation you stated. However, a different version of the line of reasoning seems to? How does one help another person become happier if helping another person become happier is the only way to become happier? One helps them help someone else to help someone to… ? Or, I suppose if it is possible to be more or less unhappy for other reasons, so by helping another person become less unhappy (though not yet happy) one could thereby become happy?
- mupuff1234 2y agoI can't fathom how anyone thought this was a sensible thing. It's so bad I have to wonder if there's a different angle here, maybe they think that releasing something so terribly bad will make it easier when they release something less comically bad? Idk.
- brookst 2y agoFact is it is hard to make a great product. I don’t think we need complex conspiracy theories to explain bad ones. Odds are the product team cherry picked conversations for internal demos, their management wanted press and promotions, KPIs were around engagement and not quality, and nobody had perspective and authority to say “this sucks, we can’t release it”.
- rsynnott 2y agoIt’s not hard to identify, before launching, “this is unacceptably bad”. Like, I think this is actually quite a bit worse than Microsoft Tay, which others mentioned, as an example of “how could you possibly launch this”? Tay was reckless, and you’ve got to assume that they knew there was at least a change it could work out as it did, but in this case the product _as launched_ was clearly unacceptable; it didn’t require input from 4chan to make it so. I mean, just very basic QA, having someone talk to the damn things for a bit, would’ve shown the problem, in this case.
- TheNewsIsHere 2y agoI’ve always believed Microsoft Tay was a really idealistic, genuine, if bombastically naïve situation.
- brookst 2y agoI wish I agreed. But I have seen many products launched where only true believers who internalized the limitations were involved in testing. Yes, a good org with self-critical people and an independent red team and execs who cared about quality would not have made this blunder. Getting that org stuff right sounds easy but is nigh impossible if it doesn’t align to the company culture.
- ErikAugust 2y agoI wonder if their product people get their ideas from Hacker News. Just the other day, a top thread comment was about an idea for a social network where you would receive tons of engagement from bots. Coincidentally (or not), Meta actually took a first step in that direction days later.
- komaromy 2y agoBig tech would never ship anything that fast.
- Semaphor 2y agoWasn’t that a comment about an article regarding the use of AI by Meta?
- debugnik 2y agoSorry for breaking the fantasy, but this is from TFA: > The company had first introduced these AI-powered profiles in September 2023 but killed off most of them by summer 2024. Also, they're deleting these profiles, which is a step away, not towards that idea. Although they're apparently leaving in some feature to make your own AI-profiles.
- paddw 2y agoNo one gets promoted for suggesting not doing something.
- nthingtohide 2y agoA mistake on par with Netflix getting into mobile games. With advent of 5G, nobody plays candy crush saga or such games.
- jedberg 2y agoYou must be kidding. Look around on public transport. If someone isn't reading something on their phone, they're almost certainly playing some candy crush like game. Apple games is doing well. Netflix failed because they didn't make games people are interested in, not because people don't want those types of games.
- nthingtohide 2y agoI think most people just scroll instagram or tiktok. Because of lack of short content format earlier, mobile games became popular. That and lack of 5g infra. Both problems have been solved.
- scarface_74 2y agoYou didn’t need 5G to stream video. You could stream video well on “3.5G” with the enhanced GSM protocols
- YurgenJurgensen 2y agoYou kind-of do need 5G to stream video on public transport where every one of the other thousand people on the line is also trying to stream video. Even LTE suffers from fundamental contention issues once you get to Tokyo/New York/London metro levels of device density.
- ghfhghg 2y agoI haven't experienced that issue in Tokyo
- scarface_74 2y ago
- deergomoo 2y agoI feel like there must be some sort of disassociation that kicks in when you spend long enough in the upper echelons of these gargantuan corporations. It's almost like spending long enough dealing with abstractions like MAUs, DAUs, and engagement metrics make you forget that actually, at the bottom of it all are real humans using your product.
- TeMPOraL 2y agoModern entrepreneurship is basically gradient descent. You try to predict what action will yield you more profits, you do that action, rinse, repeat. It's a completely abstract process.
- Mistletoe 2y agoI’m imagining it would have gone like the meme where the dissenting opinion person is thrown out the window. They don’t have any other ideas left, so they have to do stupid stuff like this to appeal to people that call themselves investors.
- jedberg 2y agoHaving worked at a similarly gargantuan and dysfunctional company, I can tell you exactly how this went down. Someone had this idea for AI profiles. They speced the product and took it to engineering. The engineers had a good laugh at how preposterous it is, but then remembered that they get paid a ton of money to do what they're told, and will get promos and bonuses for launching regardless of the outcome. It all stems from promo culture -- it doesn't matter what you build, as long as it ships.
- stephen_cagle 2y agoSeconded. It is difficult to understate how pervasive and dysfunctional promo based development is at some of these behemoths (Google from my experience, but I hear Meta is similar). Nothing else matters as long as what you are doing correctly fits in your promo packet.
- shipscode 2y agoThe best (worst?) part is when the engineers actively overlooked actual production bugs and security concerns to do this work.
- jedberg 2y agoSadly it's really hard to get a promo fixing bugs or optimizing code.
- Groxx 2y agoShip it quick, ramp up some high profile users that don't actually care much about what you're offering, and jump to the next project before anyone notices the problems. Works every time.
- loeg 2y agoThere are 10,000s of SWEs at Facebook and this project was at most a handful of SWEs. (As stupid as it is, it did not significantly detract from prod bugs.)
- 2y ago
- diego_sandoval 2y agoNext month no one is going to remember this anyway, so they don't lose much by trying.
- eastbound 2y agoWe always underestimate how many bots made the pre-2022 social apps were. 1. To make you feel like there is activity. How would you simulate activity when you have no customer to start with? I suspect Youtube threw subscribers at you just to get you addicted to video production (the only hint that my Youtube subscribers were real, was that people recognized me on the street). And guess who’s mostly talking to you on Tindr. 2. For politics and geopolitics influence. Maybe Russia pays half the bots that sway opinions on Instagram. Maybe it’s China or even Europe, and USA probably has farm bots too for Europe, to ensure the general population is swayed towards liking USA. 3. Just for business, marketing agencies sometimes suggest to create 6x LinkedIn profiles per employees to farm customers. Facebook doing it in-house and officially is just legalizing the bribe between union leaders and CEOs.
- 1659447091 2y agoThe only thing they got wrong was the strong stereotyping. But the idea from an investor or exec position is brilliant. Why bother with all the parsing for user's interactions through clicks, likes, replies etc when you can have them engage a bot. Simply have your users give you all the info you need to serve better ads; while selling companies advanced profiles on people? User profiles built by them engaging the bot and the bot slowly nudging more and more info out of them?
- ChrisMarshallNY 2y agoMy impression, from seeing some of these "great ideas," coming from the modern Tech industry, is that there really are no adults in the room; including the funders. So many of these seven-to-nine-figure disasters are of the "Who thought this was a good idea?" class. They are often things that seem like they might appeal to folks that have never really spent much time, with their sleeves rolled up, actually delivering a product, then supporting it, after it's out there.
- pianoben 2y agoWe're "all in on AI" at my job, and lots of people are drinking the koolaid. I regularly see design docs that are almost entirely written by ChatGPT, code implementing those designs written by copilot/cursor/chatgpt, and reviews done by the same. There are features deployed for which practically no human consideration was given. I'd be very willing to believe something like this happens at Facebook, too.
- honkycat 2y agoAn annoying person with a terrible idea is a powerful force at a tech company. Nobody wants to deal with them, so they just kinda get what they want as long as they are mostly harmless.