you are viewing a single comment's thread
view the rest of the comments
[–] 64 points 1 day ago (21 children)

Jesus, how far has toms hardware fallen that they pick up trumps super intelligence bullshit

  • source
  • hideshow 21 child comments
  • It’s been irritating and frustrating to watch; gamers nexus is honestly a way better source these days - not to mention, imo they are unironically one of the absolute best investigative journalists out there at this point. They’ve broken a number of very deep-dive stories, some of which have turned into big deals in the wider news media landscape (see: Intel gen 13/14 CPUs).

  • source
  • parent
  • [–] 3 points 1 day ago* (last edited 1 day ago) (1 child)

    Artificial superintelligence is a specific thing that doesn't exist yet. A ban on it would be a cautionary measure in case of future advancements. A ban on artificial intelligence writ large would arguably ban my high school project that played Tic-Tac-Toe optimally.

    It's probably important to note that ASI (and its more conservative form, artificial general intelligence or AGI) are goals, which is to say that they describe the result and not the technology used to get there. LLMs are a specific technology that I'm pretty convinced will never achieve AGI, let alone ASI. Specific-purpose AI has been around for decades (e.g. Deep Blue, which was world-class at chess in the '90s), and LLMs aren't really any different; their specific purpose is completing sentences, though, and that tricks a lot of people into thinking that they can actually perform metacognition in the way that an AGI would.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 8 points 16 hours ago

    LLM is a real engineering term.

    AI is a broad umbrella that marketers have stretched past meaning.

    AGI and SI are hypotheses about future systems with no agreed definitions, tests, or measurement, and legislation written against undefined, non-existent categories can't be enforced, only performed.

  • source
  • parent
  • [–] 1 point 17 hours ago* (5 children)

    Exactly, how do we ban a word that we just made up?

    We have LLMs, SI doesn't exist, they are not even AI.

    LLMs run a wide spectrum of types and capabilities already and most people don't even know the basics, but none of them are SI/AI.

    Should be ban local open models? (They can do "forbidden things" like copy software. They also do a ton of useful stuff, cheaply.)

    How about only government and corporate approved models with guardrails? (Think what they tell you, and sell your data? Pay forever!).

  • source
  • parent
  • hideshow 5 child comments
  • [–] 0 points 14 hours ago* (4 children)

    Should be ban local open models?

    what part of fuck ai didn't you understand when you walked in here?

    How about only government and corporate approved models with guardrails?

    No. How about not building jack SHIT on stolen data? how about not driving up the costs of ram, cpus, storage, water, power, fucking everything?

    FUCK AI. even your cutesy AI-lite flavours

  • source
  • parent
  • hideshow 4 child comments
  • [–] -1 points 13 hours ago (3 children)

    This sounds like "ban the Internet" in 1995.

    Sure push back against the hyper scalers, spread the warnings of social change (it's going to happen, we couldn't stop spam, and shortened attention spans either), but reactionary bans and hyperbolic name calling will only help the tech influencers fearmonger in their own legislation that benefits them not the people.

  • source
  • parent
  • hideshow 3 child comments
  • [+] 1 point 10 hours ago (3 children)
  • [–] -2 points 9 hours ago (1 child)

    The all caps were a nice value add.

    Well, no point having a discussion with fanatics. Enjoy fighting the steam engine John Henry. Meanwhile, some of us will at least try to make sure the benefits of the new tech go to the people and not the top 1% while you scream at the sky.

  • source
  • parent
  • hideshow 1 child comment
  • load more comments (-1 replies)
  • [–] 0 points 1 day ago* (9 children)

    Asi is an actual industry term. I haven't read the article but it is possible they do mean the specific asi rather than agi.

    Ed: If you read it they specifically mean asi not agi aka "ai" and they use the term ai and asi correctly throughout.

    https://www.tomshardware.com/tech-industry/artificial-intelligence/sanders-proposes-20-year-prison-sentence-for-ai-devs-who-plow-ahead-with-artificial-superintelligence-plans-penalty-on-par-with-illegally-developing-rogue-nuclear-weapons

  • source
  • parent
  • hideshow 9 child comments
  • [–] 2 points 20 hours ago (8 children)

    Dude, you don't know what the hell you're talking about. The thing called "AI" today is not AGI (Artificial General Intelligence).

  • source
  • parent
  • hideshow 8 child comments
  • [–] -1 points 19 hours ago (7 children)

    Perfect so we agree they're not the same terms then regardless of your dramatic starter language.

    Also no one said it was, I specifically said they're not the same thing and yes indeed agi is what most companies are after when they say ai.

  • source
  • parent
  • hideshow 7 child comments
  • [–] 2 points 10 hours ago (6 children)

    yes indeed agi is what most companies are after when they say ai.

    In the past, yes. Right now? No. The current models are not attempting to create AGI. They're trying to create a product they can market easily. Honestly, AGI probably would have marketing issues that the current models do not have. The current models are essentially statistical prediction models, not intelligence. They don't reason (despite the industry using that term for marketing). They can't understand something that is totally outside their training data. AGI could.

  • source
  • parent
  • hideshow 6 child comments
  • [–] 0 points 8 hours ago

    It's a bit silly to say they don't reason.

    Here's some research where they uncover reasoning circuits in some large transformer models https://github.com/guanzhehong/prop-logic-transformer-circuit ... reasoning has always been just math at th end of the day

    And what about alpha go and all that monte carlo tree search stuff, id argue that they can reason far better than humans in these setups

    They can't understand something...outside their training data

    Idk how easy this is for the the large models, but certainly for classic ML, testing for performance on data outside the training data is 100% part of the development process. (Read up on test/train/validation datasets)

    There is also in context learning, which gives rise to zero shot task completion for never trained on tasks.

  • source
  • parent
  • [–] 0 points 10 hours ago (4 children)

    Right now yes. They're attempting to create the architecture to build an agi. You're being tedious at this point.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 2 points 10 hours ago (3 children)

    The current industry? No, they aren't. Academia? Maybe. The current industry is not out to do something like that. They're thinking about how they can get more investor money next quarter.

  • source
  • parent
  • hideshow 3 child comments
  • [–] 1 point 9 hours ago (2 children)

    Adorable. What are they aiming for exactly? Be specific.

  • source
  • parent
  • hideshow 2 child comments
  • [–] 1 point 8 hours ago (1 child)

    I already said. Whatever makes the most money next quarter. It's not some amazing end goal they're working towards. It's whatever they con convince investors to put money into. Almost always the marketing is not about AGI. The marketing is how they're going to replace workers. You don't need AGI for this.

    It's adorable that you think it's something more intelligent than this.

  • source
  • parent
  • hideshow 1 child comment