“These re-interviews are consequential,” a leaked email seen by 404 Media reads.
    • CompactFlax@discuss.tchncs.deEnglish
      272·
      2 days ago

      That’s not enshittification; enshittification is (to simplify) making a product worse to increase profitability.

      In this case the CEO is conducting loyalty tests. It’s more of an HR issue. Groupthink.

      • 4am@lemmy.zipEnglish
        30·
        2 days ago

        Cramping AI into everything, including your “privacy” product, is enshittification.

        It’s just not really in the same form as capturing market share and then paywalling your captive market to achieve rent seeking.

        In this case you aren’t capturing rent you are capturing profiles, which in turn you will rent out to marketers.

        Proton will betray everyone. I’m glad I stayed away.

        • FlashMobOfOne@lemmy.worldEnglish
          512·
          1 day ago

          Cramping AI into everything, including your “privacy” product, is enshittification.

          That isn’t necessarily true. AI in and of itself isn’t shit. It’s the training data and the ethics of how that data was acquired that determines whether or not it is shit, and there are a ton of creators who’ve made local models trained only on their own data in order to make their work more efficient.

          Most AI users aren’t ethical in that way, but I’d trust Proton before I’d trust a frontier AI developer, if for no other reason than the laws they’re forced to operate under favor privacy. The Swiss are great like that.

          • pinball_wizard@lemmy.zipEnglish
            2·
            6 hours ago

            AI in and of itself isn’t shit.

            Sure. But AI is shit for Privacy.

            Gambling randomizers have no place in a toolchain being sold as a privacy tool.

            Proton’s leadership have lost their minds, or taken a bribe to intentionally destroy the company.

          • JustEnoughDucks@feddit.nlEnglish
            2·
            15 hours ago

            AI in and of itself isn’t shit. It’s the training data and the ethics of how that data was acquired that determines whether or not it is shit,

            And adittionally you know, the massive amounts of energy, destruction of local computing accessibility, the mass surveillance built into every chatbot, directly supporting and enabling of fascist regimes, the water stealing and toxifying, destruction of local communities, and not last or least, the Observed decline in brain function (also not limited to AI, but many digital tools, sadly).

            The only models that skip this are locally trained and run models on solar power.

            Of which it is maybe what? 0.01% of the AI use of the world? Less, if we are being completely honest?

            Proton itself uses Mistral, OpenHands, OPMO, and Qwen under the hood. Again, not ethical or avoiding any of the above problems at all.

          • badgermurphy@lemmy.worldEnglish
            7·
            1 day ago

            I guess I don’t see a good reason to trust either, and setting up a scenario where those are being compared like I must is a false dichotomy.