AI crap
drewdevault.com

There is a machine learning bubble, but the technology is here to stay. Once the bubble pops, the world will be changed by machine learning. But it will probably be crappier, not better.

What will happen to AI is boring old capitalism. Its staying power will come in the form of replacing competent, expensive humans with crappy, cheap robots.

AI is defined by aggressive capitalism. The hype bubble has been engineered by investors and capitalists dumping money into it, and the returns they expect on that investment are going to come out of your pocket. The singularity is not coming, but the most realistic promises of AI are going to make the world worse. The AI revolution is here, and I don’t really like it.

    • botengang@feddit.de
      4·
      1 year ago

      which previously failed since ads and SoC were the driver of the Web, not information.

      Can you elaborate on why you think the ads wouldn’t sneak in again? The semantic web is a fantastic concept, but I don’t immediately see the AI connection. AI doesn’t magically pay for authored content and there is still an incentive to somehow get ads into LLM answers.

        • xavier666@lemm.eeEnglish
          4·
          1 year ago

          You can run a LLM at home on your own PC. You tell it what you want and it goes to search the net for you.

          Unless it’s open-source and connected to a proper crownsourced dataset, hosted on a paid server managed by a community instead of a big corporation, I don’t see how ads are NOT getting in.

        • botengang@feddit.de
          2·
          1 year ago

          Thank you very much. My concern is rather in the direction of inserting ads or “promotional information” into the training material, much like SEO plagues search today. If the info is from the web it can still be malicious, even if you run your own LLM.

            • david@feddit.uk
              1·
              1 year ago

              I don’t know why you want to use an AI to purchase goods and learn about products. That’s what the current www is really really strong at. Lots of people are spending an awful lot of money to make that information really easy to discover, and popular search engines definitely prioritise that information.

              Also, if an AI is to give you price and product information it’s going to have to be reading live web pages, which will of course be full of ads. SEO will become AIO/LLMO. There is no end to the time and money advertisers are prepared to pour into getting products in front of users. The irony is that you seem to want to view products and you have this weird perspective where you’re keen to avoid ads for products so that you can view marketing information about products without the ads.

              It’s already fairly hard to tell without knowing some good websites or reading through to conclusions and using some common sense whether a review website is honest or biased. I don’t know why you think an AI with access to the Internet will filter out fake reviews and content crafted to lead you to specific products over others.

              Also, downloading and configuring your own AI is unlikely to be the way the “AI revolution” comes. Amazon, Google, Microsoft, Apple and other mega corporations will be funding the “AI revolution” and will not sit idly by allowing their kingdoms to crumble.

              The number of people who will be saved from the corporations that run the online world by open source grass roots AI will be smaller than the number of people who are saved by Linux from proprietory products and SAAS.

              Yeah, everyone will get used to using an AI to interact with the web, but it will be freely supplied by a corporation, and I PROMISE you the enshitification of AI has been long planned before we even reach step one of making it awesome for the masses.

                • david@feddit.uk
                  1·
                  1 year ago

                  You know that a LLM is a statistical word prediction thing, no? That LLMs “hallucinate”. That this is an inevitable consequence of how they work. They’re designed to take in a context and then sound human, or sound formal, or sound like an excellent programmer, or sound like a lawyer, but there’s no particular reason why the content that they present to you would be accurate. It’s just that their training data contains an awful lot of accurate data which has a surprisingly large amount of commonality of meaning.

                  You say that the current crop of LLMs are good at Wikipedia style questions, but that’s because their authors have trained them with some of the most reliable and easy to verify information on the Web. A lot of that is Wikipedia style stuff. That’s it’s core knowledge, what it grew up reading, the yardstick by which it was judged. And yet it still goes off on inaccurate tangents because there’s nothing inherently accurate about statistically predicting the next word based on your training and the context and content of the prompt.

                  Yes, LLMs sound like they understand your prompt and are very knowledgeable, but the output is fundamentally not a fact-based thing, it’s a synthesized thing, engineered to sound like its training data.