• 1 Post
  • 12 Comments
Joined 1 year ago
cake
Cake day: June 12th, 2023

help-circle

  • The guy who leads this group is extremely vocal (almost weirdly so) about white privilege and systemic racism. He is also white. It’s true that many AI models have white-bias. The reasons for this are multi-faceted. Our datasets are grossly imbalanced against racial minorities. I also think I understand that for some darker-skinned races, it is more difficult for the model to extract relevant features from the shitty Flickr photos they scrape for these models.

    That said, injecting words into the users prompt to force the model to generate minorities more often is an extremely naive approach. Kind of like if Google added “reddit” to all searches just because it worked for some specific test cases, but ignoring that you now no longer get any site except reddit. Probably the solution here looks like paying a lot of money for high quality datasets as well as investing in user education and more AI explainability of these tools.


  • Prophet@lemmy.worldtoPeople Twitter@sh.itjust.worksThe dream
    link
    fedilink
    English
    arrow-up
    2
    ·
    11 months ago

    I agree, in the context of the tweet, that purchase history is enough to build a working product that roughly meets user requirements (at least in terms of predicting consumed items). This assumes you can find enough purchase history for a given user. Even then, I have doubts about how robust such a strategy is. The sparsity in your dataset for certain items means you will either a.) be forced to remove those items from your prediction service or b.) frustrate your users with heavy prediction bias. Some items also simply won’t work in this system - maybe the user only eats hotdogs in the summer. Maybe they only buy eggs with brownie mix. There will be many dependencies you are required to model to get a system like this working, and I don’t believe there is any single model powerful enough to do this by itself. Directly quantifying the user’s pantry via vision seems easy in comparison.


  • Prophet@lemmy.worldtoPeople Twitter@sh.itjust.worksThe dream
    link
    fedilink
    English
    arrow-up
    5
    ·
    11 months ago

    Also quite difficult from a vision perspective. Tons of potential object classes, objects with no class (e.g., leftovers, homemade things), potential obfuscation if you are monitoring the refrigerator/cabinets. If the object is in a container, how do you measure the volume remaining of that substance? This is just scratching the surface I imagine. These problems individually are maybe not crazy challenging but they are quite hard all together.


  • I also really like these things. I read the article, and I got the sense that their rationalization for picking BG3 as GOTY was based exclusively on the actual game design, and before I say anything else, I agree that the design is fantastic. But, I would love to see Larian get more credit for the transparency during development, their commitment to delivering an experience of the highest quality, and the fairness and respect they give to their player base.

    Whether or not a person likes games like BG3 or thinks it deserves GOTY, this stuff alone should be the bar for all games, not the scummy greedy practices we’ve become so used to seeing. Thank you Larian Studios!


  • Prophet@lemmy.worldtocats@lemmy.worldCats understand 'naughty'
    link
    fedilink
    English
    arrow-up
    56
    arrow-down
    1
    ·
    1 year ago

    I left out a crockpot of mostly eaten mac and cheese on the counter. I was on the couch half-asleep when I heard my keys (which were next to this crockpot) jingle. I didn’t say anything, I just turned my head and saw my cat running for cover as if it had just tripped the alarms during a heist gone wrong. How do I interpret this in any way other than my cat knowing what it was doing was naughty?



  • Microsoft didn’t “absorb” open ai, they have a partnership where Microsoft pays assloads of money to sustain openai so that Google doesn’t get it. Ironically, this might be considered “long term thinking” but I wonder how long shareholders will tolerate such a hit to the books. There is supposed to be a profit sharing model here eventually (up to a certain point) but Microsoft isn’t getting chatgpt, otherwise bing would have replaced chatgpt. I have to wonder if, by the time chatgpt is profitable, if there will already be better models produced by other groups (maybe even open source), especially given the pace of AI innovation. I would not be surprised if this was a net loss for MS. GPT is amazing but it has numerous drawbacks at the moment. I admit that, if they figure things out quickly, this could be a huge win for them. I would go so far as to say that this is not anti consumer at all and is exactly how the free market is supposed to work.

    As for Facebook, the only data you need is that the younger generations think it’s for boomers and don’t use it. I’m a little older and (to your credit) I check in about once a month. I know that meta has a very powerful user data harvesting business (arguably more valuable than Facebook), but Facebook’s user engagement will continue to slide if they can’t capture younger users and keep millennials and gen x users on the platform. This devalues their ability to make money from ads directly, and again, they did this to themselves by destroying their reputation for short term gains. They will eventually become like Yahoo! or AOL, both of which have almost zero brand value.


  • I’m not sure it’s entirely accurate to say these companies aren’t destroying themselves though. Are they just going to explode and die all at once? Probably not, but they will likely fade to obscurity like IBM or HP (two powerhouses of the last century). I agree that exploiting customers is how they make money hand over foot (and we just roll over for it) but the point is to make the largest possible short term gains, not to maximize profit. It’s important to maximize short term gains because it makes big shareholders happy, and the shareholders (e.g., the CEO and the board) want to enrich themselves. The issue with optimizing for short term gains is that you miss out on the dividends of long term effort, which is usually significantly greater.

    Something I think about occasionally is how it is that a no-name startup beat the likes of Google, MS, Facebook, etc to chatgpt. Chatgpt is the single greatest innovation in search in almost 3 decades. Google’s whole business relies on users needing Google’s search platform to find information. Google gets to place ads here, and that makes up the largest part of their revenue, but chatgpt threatens to upend that whole business. There is the potential for a whole new generation of advertisement technology to be baked into chatgpt that delivers an unprecedented level of ad targeting. In case you need a translation, that is massive $$$$$$$$, because advertisers want their ads to be placed in front of people who will actually buy the product (and they will pay a premium for this!), not the spray and pray strategy you see today.

    So yes, in a way, Google and other companies that rely on simply extracting wealth rather than innovating/building wealth risk losing billions of dollars and eventually fading to irrelevance. I really think Facebook has passed the point of no return already in this regard, and has allowed numerous social media sites to steal market share very easily.


  • You can see how those two things are a little different though, right?

    No, not really. Contrary to your point, Bethesda has worked quite closely with Xbox a number of times (especially back in the oblivion days) and Sony has never been interested in Bethesda’s ideas about games (support for Skyrim was abysmal on PlayStation and mods on PS3/4 were a joke).

    Is MS a huge jerk for yanking starfield out of the hands of the majority of console gamers? Yeah totally, but Sony is also a huge jerk (and has been) for a long time when it comes to negotiating exclusivity deals, which they have been able to do because they are the number 1 console. It’s really not hard to extrapolate how much leverage Sony has over the industry when you see that they have sold 75% more consoles than xbox (35 vs 20 million units sold PS5/XS). I believe the previous gen was even worse. The outcry over this would have been much smaller if the roles were reversed, because it would have just been business as usual for every gamer.


  • Since the starfield exclusivity thing started, this point has always stuck with me: PlayStation owners buy PlayStation because of the expectation that they will get the best exclusives (and even most other games first). It was so bizarre to see them so brazenly attack Xbox over making starfield exclusive. They couldn’t see that they were beneficiaries of these same tactics for so long that they just accepted it as “the way it is.” Logically, why would you ever buy an Xbox if PlayStation gets better exclusives and the other great games first? No one should be surprised when TES6 is Xbox/PC exclusive.