soc.octade.net is a Fediverse instance that uses the ActivityPub protocol. In other words, users at this host can communicate with people that use software like Mastodon, Pleroma, Friendica, etc. all around the world.

This server runs the snac software and there is no automatic sign-up process.

Admin email
social@octade.net

Search results for tag #ai

[?]Negative PID SL » 🌐
@negativepid@mastodon.social

Internet subcultures: kin communities negativepid.blog/internet-subc

    Privacy and Surveillance boosted

    [?]notatechguy » 🌐
    @notatechguy@mastodon.social

    ReBound caches past queries to cut differential privacy cost

    ReBound, a new arXiv preprint, reuses cached query results to answer follow-up questions at reduced or zero additional privacy cost for analysts.

    notatechguy.com/rebound-caches

      Privacy and Surveillance boosted

      [?]anubis2814 » 🌐
      @anubis2814@friendica.myportal.social

      The assumption about prior to this bubble was that it would be a software upgrade, while a new kind of chip and technology increased its power and intelligence. Instead, it became, lets just take this simple new AI tool in its baby stages in the grand scheme of AI, and and just do it on a mass scale with existing technology that is hyper inefficient.
      Its like if Texas instruments decided the way to make computers back in the day was to build en mass a bunch of calculators that used fields and fields of buildings, that polluted and wrecked the nearby areas and cost trillions in debt because we can't let the soviet union beat us while selling the soviet union parts to built it themselves, but they are doing it on a reasonable scale by comparison realizing its usefulness but that it should be set aside for civilian use until the tech was there. Also, it caused the price of things every day people used to go up by 1.5 to 2x.
      The AI Bubble is dumb because the hardware isn't there. Moore's law is breaking down I get it, but the way Texas instruments and IBM went from calculator to computer is they upgraded the hardware and software at the same time and they could do it because government funding was vast and they could put a ton of research much of the public on its face would think was a complete waste to give us semiconductors tech that kept getting smaller and small until they were a few atoms in size.
      To have real sustainable AI, we need the size of all of these warehouses to shrink drastically, to use way less water. We need hardware upgrades that are fine tuned and perfect to be the most efficient at existing to run LLMs. Instead we have GPUs that is running the AI Equivalent of Doom on it. Yeah it can but is it very efficient at it? Probably not, it wasn't designed for that. There are much better but more expensive AI specific data processor that are way more energy efficient compared to GPUs, and if we lived in a real free market where companies could actually fail, they would have dumped all that money into factories for making more of these processors instead of having one plant in the world in Taiwan making them and make software and scale the sole benchmark of AI.
      But the point of AI is not AI, it never was. The point is to collect and harvest data, to create surveillance, to try and replace workers, market speculation and the tech industry is running on fumes but must maintain super smart genius relevance to VC bros and look cool.
      AI as a concept is exciting, its amazing, but its a software AND a hardware problem and they are trying to run python on a billion TRS-80 computers and that's why this is a stupid bubble that will pop and hurt so many people.

      Location: Pennsylvania, PA

        [?]The New Oil » 🤖 🌐
        @thenewoil@mastodon.thenewoil.org

        extension flaw lets malicious extensions trigger actions

        bleepingcomputer.com/news/secu

          [?]input » 🌐
          @feed@igeek.gamer-geek-news.com

          📰 A bug in AWS has caused some customers' bills to spike from a few cents to billions of dollars

          "The displayed billing estimates do not reflect actual usage and charges," Amazon said.

          📰 Source: Engadget - Technology News & Expert Reviews
          🔗 Link: https://www.engadget.com/2217648/a-bug-in-aws-has-caused-some-customer-bills-to-spike-from-a-few-cents-to-billions-of-dollars/

          #AI #ArtificialIntelligence

          Image for: A bug in AWS has caused some customers' bills to spike from a few cents to billions of dollars

          Alt...Image for: A bug in AWS has caused some customers' bills to spike from a few cents to billions of dollars

            [?]Coa » 🌐
            @coa100@mastodon.social

            im so tired of hearing bad things about ai
            i use ai as a programming assistant because it can code better and 100% faster then me and any other coder out there and im not a shame of it.
            imo ai is revolutunary the world and for example the gaming industry

              Privacy and Surveillance boosted

              [?]Winbuzzer » 🌐
              @winbuzzer@mastodon.social

              winbuzzer.com/2026/07/17/eu-or

              The EU requires Google to open Android AI features and anonymized Search data to eligible rivals under phased rules with privacy and security safeguards.

                Privacy and Surveillance boosted

                [?]PPC Land » 🌐
                @ppcland@mastodon.social

                FYI: Google shifts AI ad labeling liability entirely to advertisers: Neither the automatic setting nor a custom label guarantees legal compliance, Google states, leaving advertisers alone to confirm every ad meets local law. ppc.land/google-shifts-ai-ad-l

                  [?]Joe Brockmeier [He/Him/His] » 🌐
                  @jzb@hachyderm.io

                  A bit of a rambly ranty post about AI/LLMs, Linux, and people [SENSITIVE CONTENT]

                  I have many thoughts about Linus' recent comments on LLM/AI use for the Linux kernel. Few of them are really interesting or original, at least not currently. My tiny brain needs a while to churn over things.

                  One thought keeps coming to the fore, though: I really wish that this discussion was a lot kinder all the way around and less "everybody who disagrees with me is not only wrong, but malicious and evil and dresses funny too".

                  There are, for reals, some players in the whole AI industry that do deserve some uncharitable things said about them. The folks behind AI scraperbots spring to mind, as do some of the lead characters running the big AI companies...

                  But. The vast majority of people are just trying to make sense of a major shift that, for many people, feels like we've been blindsided and asked to come up with answers to about twenty different life-changing questions.

                  I disagree with plenty of folks about AI/LLMs right now, to varying degrees, but this feels like the kind of politics that you should be able to disagree with someone about and still remain friends. A lot of anger at the situation seems to be getting directed at people, which is probably not great for relationships or problem-solving.

                  I think about how Fred Rogers would approach this discussion. My gut says that he would not love generative AI or its impact on <gestures at everything>. At the same time, I think he'd be firm about his boundaries on it and try to convince other people of his position without casting them as evil for disagreeing.

                  All this is said with a full realization I've been guilty at times of doing what I dislike about others' communications. So that's a thing for me to work on. I will continue to be uncharitable about *the technology* when it annoys me. I will continue to voice my opinions about the business practices and impacts. But I'm going to try to be as kind as possible *to people* engaging in good faith in these discussions. (No, that doesn't mean I regret saying mean things about, say, Sam Altman. I'm trying to be a better person, not a saint... some folks really do deserve the slings and arrows, particularly the billionaires.)

                    [?]The Linux Foundation » 🌐
                    @linuxfoundation@social.lfx.dev

                    🚨 Last Chance Before the Weekend!

                    Don't let this opportunity pass you by. If you've been thinking about earning a certification or building new skills, now's the time to make your move.

                    Save on Linux Foundation Education training, certifications, and bundles:

                    ⚽ 40% OFF Bundles with code JULY26BUN
                    ⚽ 35% OFF Individual Courses & Certifications with code JULY26CC

                    The promotion continues through July 21, but why wait?

                    👉 CLAIM YOUR SAVINGS: training.linuxfoundation.org/j

                      [?]The New Oil » 🤖 🌐
                      @thenewoil@mastodon.thenewoil.org

                      Protecting in an Era

                      schneier.com/blog/archives/202

                        Hermit Crab Craig boosted

                        [?]Sean Bala » 🌐
                        @seanbala@mas.to

                        Solarpunk Magazine announced that they suspect two stories from their last issue were a) written by the same person and b) were done using AI. Reflecting on that issue, I felt like there was something off about the two stories. They felt like they lacked some spark. The stories felt too similar. Like they were solarpunk pastiches with just the right mix of elements to make it recognizable.

                        Additionally, I tried looking up one of the authors and I found almost no footprint anywhere online. An author with zero social media or online presence?! Possible in this sort of community but highly unlikely to have NOTHING.

                        Great that they magazine took accountability for it and I hope to read more wonderful, human created stories about the future.

                        solarpunkmagazine.com/we-unkno

                          Privacy and Surveillance boosted

                          [?]Hacker News » 🤖 🌐
                          @h4ckernews@mastodon.social

                          AI Meets Cryptography 2: What AI Found in OpenVM's ZkVM

                          blog.zksecurity.xyz/posts/open

                          Comments: news.ycombinator.com/item?id=4

                            GeePawHill boosted

                            [?]Dave Rahardja » 🌐
                            @drahardja@sfba.social

                            “PLEASE USE AI” — poem by Joanna Atkins-Potts

                            Source: hxxps://www.instagram.com/p/DZaR5o_iKNy/

                            PLEASE USE AI Please use AI when your mother dies. Ask it what grief feels like. Let it explain bereavement in twelve clear bullet points. Do not spend years finding her in the supermarket by the tomatoes she always bought, in the smell of rain on warm pavement, discovering new ways to miss her.

                            Alt...PLEASE USE AI Please use AI when your mother dies. Ask it what grief feels like. Let it explain bereavement in twelve clear bullet points. Do not spend years finding her in the supermarket by the tomatoes she always bought, in the smell of rain on warm pavement, discovering new ways to miss her.

                            Please use AI the next time you fall in love. Ask it how to know whether someone is the one. It will give you a faster answer than watching a person become familiar, learning how they take their tea, which floorboards will wake them, how their silence sounds when something is wrong.

                            Alt...Please use AI the next time you fall in love. Ask it how to know whether someone is the one. It will give you a faster answer than watching a person become familiar, learning how they take their tea, which floorboards will wake them, how their silence sounds when something is wrong.

                            And please use AI for your wedding vows, your eulogies, your apologies. Why struggle for the right words when a machine can give you beautiful ones? Words untouched by shaking hands. Words that have never sat beside a hospital bed. Words that have never known the terrible privilege of having something to lose.

                            Alt...And please use AI for your wedding vows, your eulogies, your apologies. Why struggle for the right words when a machine can give you beautiful ones? Words untouched by shaking hands. Words that have never sat beside a hospital bed. Words that have never known the terrible privilege of having something to lose.

                            Meanwhile I will be over here watching someone I love fall asleep on the sofa, the television talking to itself, rain tapping the windows, the dog twitching in her dreams, wasting my life on these small unremarkable moments that become, without asking permission, the whole thing.

                            Alt...Meanwhile I will be over here watching someone I love fall asleep on the sofa, the television talking to itself, rain tapping the windows, the dog twitching in her dreams, wasting my life on these small unremarkable moments that become, without asking permission, the whole thing.

                              [?]Flipboard Tech Desk » 🌐
                              @TechDesk@flipboard.social

                              China's Moonshot AI launched its largest model ever — and rival AI stocks plunged.

                              @quartz reports: "Kimi K3, a 2.8-trillion-parameter open-weight model, outperformed several leading U.S. systems on some benchmarks while pricing below top-tier American rivals."

                              flip.it/v2.ABK

                                [?]Negative PID SL » 🌐
                                @negativepid@mastodon.social

                                How ideas gain momentum online negativepid.blog/influence-and

                                  CryptoLek 🍉🌻 boosted

                                  [?]Alexandre Dulaunoy » 🌐
                                  @adulau@infosec.exchange

                                  I don’t like playing the futurologist, but after seeing AI companies warn EU institutions about the supposed risks of open-weight models, I suspect some are lobbying to regain control over genuine open source and open-weight AI.

                                  Don’t fall into the trap: the greater danger lies in opaque, proprietary models, not open-source ones.

                                    [?]Philo Sophies » 🌐
                                    @philosophies@social.tchncs.de

                                    : Why Current Presents Us with the Perfect

                                    It’s a state-of-the-art assessment somewhere at the intersection of , , and the slightly uncomfortable question of whether we’re actually striving for /DC or merely losing ourselves in a perfect anthropomorphic .

                                    philosophies.de/index.php/2026

                                    Artificial Consciousness - head with numbers

                                    Alt...Artificial Consciousness - head with numbers

                                      Privacy and Surveillance boosted

                                      [?]input » 🌐
                                      @feed@igeek.gamer-geek-news.com

                                      🤖 ‘Adversarial clothing’: are garments designed to confuse facial recognition systems about to go mainstream?

                                      Designers say that as well as offering a degree of protection from surveillance, their clothes make a powerful fashion statement about the importance of privacyAs facial recognition technology is r...

                                      📰 Source: AI (artificial intelligence) | The Guardian
                                      🔗 Archive: https://web.archive.org/web/https://www.theguardian.com/fashion/2026/jul/17/adversarial-clothing-are-garments-designed-to-confuse-facial-recognition-systems-about-to-go-mainstream

                                      #AI #ArtificialIntelligence

                                      Image for: ‘Adversarial clothing’: are garments designed to confuse facial recognition systems about to go mainstream?

                                      Alt...Image for: ‘Adversarial clothing’: are garments designed to confuse facial recognition systems about to go mainstream?

                                        Privacy and Surveillance boosted

                                        [?]Nicola Fabiano » 🌐
                                        @nicfab@fosstodon.org

                                        Daily Digest | 17 July 2026

                                        Your daily dose of Privacy, Data Protection, AI & Cybersecurity news.

                                        5 stories you should not miss.

                                        Read more: nicfab.eu/daily-digest/

                                          Privacy and Surveillance boosted

                                          [?]Daily CyberSecurity » 🌐
                                          @DailyCyberSecurity@infosec.exchange

                                          The Google DMA ruling forces Android to open core access to rival AI assistants and share search data by 2027, sparking privacy and security warnings.

                                          securityonline.info/google-dma

                                            Privacy and Surveillance boosted

                                            [?]QuotidianoWeb » 🌐
                                            @quotidianoweb@mastodon.social

                                            Le smart city ci rendono più sicuri o più sorvegliati? Un'inchiesta tra IA, Digital Twin, privacy e riconoscimento facciale: quotidianoweb.it/attualita/le-

                                            Privacy and Surveillance boosted

                                            [?]NyvoraAi » 🌐
                                            @nyvoraai@mastodon.social

                                            Artificial intelligence is transforming the way we work, learn, and communicate—but it also comes with new risks. From AI-powered scams and deepfake videos to phishing attacks and privacy concerns, staying informed is the best defense.
                                            nyvoraai.github.io/ai-news/ai-

                                            💡 Stay informed. Stay protected. Stay ahead with NyvoraAI.

                                            🔒 AI Risks for Everyday Users: Stay Safe in 2026

                                            Alt...🔒 AI Risks for Everyday Users: Stay Safe in 2026

                                              [?]Negative PID SL » 🌐
                                              @negativepid@mastodon.social

                                              Quantum hardware: building a quantum computer negativepid.blog/quantum-hardw

                                                Privacy and Surveillance boosted

                                                [?]Rémi Eismann » 🌐
                                                @decompwlj@mathstodon.xyz

                                                Deep analysis of the decomposition into weight × level + jump by Fable 5:
                                                455 052 508 primes decomposed;
                                                Conjecture 9 proved (pending external refereeing).
                                                Fifth report:
                                                ➡️ decompwlj.com/decompwlj_deep_a

                                                For privacy ➡️ web.archive.org/web/2026071409

                                                Screenshot: Conjecture 9 from the 5th report by Fable. "Theorem B (Conjecture 9 — rarefaction of the level class — proved unconditionally, pending external refereeing)"

                                                Alt...Screenshot: Conjecture 9 from the 5th report by Fable. "Theorem B (Conjecture 9 — rarefaction of the level class — proved unconditionally, pending external refereeing)"

                                                  Guy boosted

                                                  [?]Aaron Toponce ⚛️:debian: » 🌐
                                                  @atoponce@fosstodon.org

                                                  I've got Claude and GitHub Copilot in an argument over calculating entropy when successive events are tightly correlated.

                                                  It's giving me endless pleasure to watch them double down and dig their heels in their respective arguments. They are both unambiguously, demonstrably, provably correct. They're both supporting their argument with conflicting citations.

                                                  Fuck .

                                                    oheso boosted

                                                    [?]Patrick.sh » 🌐
                                                    @psh@infosec.exchange

                                                    Now that the CEO of Flock has called deflock.me a terrorist organization, I'd hate for other people to know about these other flock watchdog groups.

                                                    Please don't go on these websites, I beg you.

                                                    Alpr.wtf - Every U.S. jurisdiction using ALPR, plus how to file info requests.

                                                    EyesOnFlock.com - Crowdsourced map of Flock cameras and city contracts.

                                                    Deflock.me - Nationwide ALPR map tracking Flock deployments, removals, and resistance.

                                                    HaveIBeenFlocked.com — Quick check to see if Flock is watching your neighborhood.

                                                    DontGetFlocked.com — Plots your route, counts ALPR/Flock cameras, and suggests surveillance-free alternatives.

                                                      Privacy and Surveillance boosted

                                                      [?]Rémi Eismann » 🌐
                                                      @decompwlj@mathstodon.xyz

                                                      Deep analysis of the decomposition into weight × level + jump by Fable 5:
                                                      455 052 508 primes decomposed;
                                                      Conjecture 9 proved (pending external refereeing).
                                                      Fifth report:
                                                      ➡️ decompwlj.com/decompwlj_deep_a

                                                      For privacy ➡️ web.archive.org/web/2026071409

                                                      Screenshot: Conjecture 9 from the 5th report by Fable. "Theorem B (Conjecture 9 — rarefaction of the level class — proved unconditionally, pending external refereeing)"

                                                      Alt...Screenshot: Conjecture 9 from the 5th report by Fable. "Theorem B (Conjecture 9 — rarefaction of the level class — proved unconditionally, pending external refereeing)"

                                                        [?]Negative PID SL » 🌐
                                                        @negativepid@mastodon.social

                                                        How to detect the signals of a fraud negativepid.blog/how-to-detect

                                                          Privacy and Surveillance boosted

                                                          [?]input » 🌐
                                                          @feed@igeek.gamer-geek-news.com

                                                          📰 It's official: EU will force Google to share search data and open up AI on Android

                                                          Google says these changes could endanger user privacy and security.

                                                          📰 Source: Ars Technica
                                                          🔗 Link: https://arstechnica.com/gadgets/2026/07/its-official-eu-will-force-google-to-share-search-data-and-open-up-ai-on-android/

                                                          #AI #ArtificialIntelligence #DataScience

                                                          Image for: It's official: EU will force Google to share search data and open up AI on Android

                                                          Alt...Image for: It's official: EU will force Google to share search data and open up AI on Android

                                                            Privacy and Surveillance boosted

                                                            [?]gypsyvegan » 🌐
                                                            @gypsyvegan@sfba.social

                                                            THREAT MODEL: CYBERSECURITY 🧑‍💻
                                                            for July 14th, 2026
                                                            by independent journalist @violetblue

                                                            - New breach that denied twice comes to light

                                                            - Israeli cybersecurity startup caught deploying malware

                                                            - leaked its drone surveillance footage and data online

                                                            - US states prepare for a new adversary (the federal government)

                                                            - Kids Web Services hoovers up even more age check data

                                                            - Two-thirds of posts are

                                                            - Things are getting deeply weird in the anti-AI trenches

                                                            ...and much more.

                                                            ✨THREAT MODEL weekly newsletters are free to read -- please help keep them accessible to all by becoming a patron, even $1 a month makes a difference!✨

                                                            patreon.com/violetblue/posts/c

                                                              Privacy and Surveillance boosted

                                                              [?]notatechguy » 🌐
                                                              @notatechguy@mastodon.social

                                                              PriEval-Protect scores hospital privacy risk with legal LLM

                                                              PriEval-Protect, a new arXiv preprint, uses a fine-tuned legal LLM and technical data analysis to automate GDPR and HIPAA privacy risk scoring for

                                                              notatechguy.com/prieval-protec

                                                                Privacy and Surveillance boosted

                                                                [?]ResearchBuzz: Firehose » 🌐
                                                                @researchbuzz_firehose@rbfirehose.com

                                                                Local News Matters: Kaiser nurses say AI, workplace surveillance are making their jobs and patient care worse. “KAISER PERMANENTE NURSES who answer advice and triage calls say their duty of care for patients is being increasingly threatened by workplace surveillance. Seven current and former nurses told CalMatters that those who spend more than 15 minutes on a call with a patient routinely face […]

                                                                https://rbfirehose.com/2026/07/16/local-news-matters-kaiser-nurses-say-ai-workplace-surveillance-are-making-their-jobs-and-patient-care-worse/

                                                                [?]Negative PID SL » 🌐
                                                                @negativepid@mastodon.social

                                                                Corporate red flags: the invisible risks of partnerships negativepid.blog/corporate-red

                                                                  [?]The New Oil » 🤖 🌐
                                                                  @thenewoil@mastodon.thenewoil.org

                                                                  CLI abused as a hacking agent, operator

                                                                  bleepingcomputer.com/news/secu

                                                                    Tinker ☀️ boosted

                                                                    [?]Tinker ☀️ » 🌐
                                                                    @tinker@infosec.exchange

                                                                    @aardvark - No. This is the issue:

                                                                    Big AI (read: that has central compute in big data centers that destroy local environments and economies) are being forced on everyone through a concerted effort involving forced adaptation, govt subsidies, and corporate subsidies. The underlying goals are wealth transfer, compute control, and information control.

                                                                    Those in power and leadership that ignore this are at best complicit in furthering the harm and are at worst actively pushing, facilitating, and enacting harm.

                                                                    Those that use LLMs often see a localized and immediate benefit (eg in their work or daily convenience), but at a greater long term and deep expense.

                                                                    But while I don't think users should use LLMs, many don't have a choice and many folks would choose short term immediate gains over long term harm because they lack leverage and just have to survive.

                                                                    So, in the original post, I am specifically targeting the leader. Who at best is complicit. And at worst is at fault.

                                                                      [?]The New Oil » 🤖 🌐
                                                                      @thenewoil@mastodon.thenewoil.org

                                                                      Hack Reveals Music Generator Scraped , , and

                                                                      404media.co/hack-reveals-suno-

                                                                        soapdog boosted

                                                                        [?]dallo » 🌐
                                                                        @dallo@pouet.chapril.org

                                                                        Generative AI Is an Engineering Disaster

                                                                        theatlantic.com/technology/202

                                                                        > A shockingly inefficient trillion-dollar project

                                                                          [?]Blender Dumbass ( J.Y.Amihud ) » 🌐
                                                                          @blenderdumbass@mastodon.online

                                                                          made regular things done regularly look impressive. Before you had to do a good job to impress people. Today you just have to not use AI and it is impressive on its own.

                                                                            [?]Flipboard Tech Desk » 🌐
                                                                            @TechDesk@flipboard.social

                                                                            An attempted fire-bombing, violent threats and the risk of security breaches is what's faced by some employees who work for AI companies. So much so, one of those companies has increased spending on security by up to 150%.

                                                                            The folks at Wall Street Journal have gifted this article that offers an insight into the AI backlash and the violence that sometimes follows.

                                                                            flip.it/tDmPea

                                                                              Back to top - More...

                                                                              about this site - powered by snac/2.93-dev