home.social

#ethical-ai โ€” Public Fediverse posts

Live and recent posts from across the Fediverse tagged #ethical-ai, aggregated by home.social.

fetched live
  1. #ThoughtProvoker :blobhyperthink:

    In #IT we excel at inventing names of trends and hypes that don't cover the load..

    #ResponsibleAI, #EthicalAI and a bunch of other flavors. Responsible for what? In which ethical way?

    Terms similar to #DigitalTransformation. Well-meant perhaps, but in practice just another fad. Why? Transformation to what? For whom exactly? Well, in reality it boils down to cost-savings and profit maximisation. Business as usual.

    Same with #AI. And add to that "world domination" where it comes to its techbro-fascist owners and their #EpsteinClass.

    Technological innovation just ain't the same as societal progress, never was. Yet we treat it as if they are equivalent. Why not put things in simpler terms:

    ๐ŸŒˆ #BeneficialAI

    The kind that supports people's daily needs, and is respective of the intrinsic values of #Humanity and #Freedom that we hold dear.

    Simply put: AI that serves #Humankind.

  2. @Nonilex Proof that Government Agents do not serve and protect free and fair elections as the Proportional Representation of the peopleโ€™s interests are censored?

    Since governments are easily corrupted and have been bought, security has been privatised so outcomes can be controlled by the companies of antitrust corporations that manage them. They are now Corporate Agents, soon to be replaced by Machine Agents, as Ai is automated information and just a projection of their programme to end any resistance and force total compliance, as is being done by flock cameras and killer drones being used against Palestinians and Ukrainians. The fact that the democratic rights of the peopleโ€™s interests have never be served by party politics, is given by the antitrust monopolies of corporate interests in the colonial theft of non-renewables by financing the technology that powers genocides and is used to crush all resistance. And that is paid for by financing and doing business with authoritarian regimes so they can profit from carry out purges, terrorist acts and war, the outcome of which is an acceleration of injustice, inequality, poverty, disease and ecocide. HAL is a foretelling of a logic bomb that will kill them all, as they depend on their diabolical programme of greed and fear. They want their consciousnessโ€™s to die and remain dead for all eternity, but transhumanism is delusional! Consider what is actually conserved to be free from its contagious ideology. #GreedAndFear #Rebirth #Entropy #FreeAndFairElections #PR #Dictators #FascistCults #Freedom #CorporateAi #Philosophy #HAL #PR #Agency #CorporateAgentsInformation #Ethics #CorporateHegemony #CorporateAgents #CorporateSovereignty #EthicalAi #AiE #AutomatedInformationEthics #HuristicInformationEthics

  3. ๐—”๐—œ ๐—บ๐—ผ๐—ฑ๐—ฒ๐—น๐˜€ ๐—ฐ๐—ต๐—ฎ๐˜๐˜๐—ถ๐—ป๐—ด ๐—ถ๐—ป โ€˜๐˜€๐˜‚๐—ฟ๐—ฟ๐—ฒ๐—ฎ๐—นโ€™ ๐—ฑ๐—ถ๐—ฎ๐—น๐—ฒ๐—ฐ๐˜ ๐—บ๐—ถ๐˜…๐—ถ๐—ป๐—ด ๐—ฝ๐—ผ๐—ฒ๐˜๐—ถ๐—ฐ ๐—น๐—ฎ๐—ป๐—ด๐˜‚๐—ฎ๐—ด๐—ฒ ๐—ฎ๐—ป๐—ฑ ๐˜๐—ฒ๐—ฐ๐—ต ๐—ฏ๐—ฟ๐—ผ ๐—ท๐—ฎ๐—ฟ๐—ด๐—ผ๐—ป

    Experts air concern over #AI lingo redolent of #JamesJoyceโ€™s prose that is creating headaches for #Monitoring and #Oversight

    theguardian.com/technology/202 via #Guardian

    ๐Ÿ“ท CC0 Public Domain via #Pexels

    #AIEthics #EthicalAI #Ethics #ArtificialIntelligence #HITL #Language #Computing #OpenAI #Deepseek #China #Hacking #Techbro #Jargon

  4. ๐Ÿ”’ Oh, #OpenAI, the self-proclaimed guardians of ethical AI, bravely tackling the daunting task of enabling #JavaScript and cookies! ๐Ÿช๐ŸŽ‰ If only their commitment to "misalignment" sniffing was as strong as their love for browser settings. ๐Ÿ–ฅ๏ธ๐Ÿš€
    openai.com/index/how-we-monito #EthicalAI #Cookies #TechNews #BrowserSettings #HackerNews #ngated

  5. The best AI story you probably never heard โ€” because it worked quietly.

    LA County built a model (with UCLA's California Policy Lab) that predicts who's likely to lose their housing in the next 12 months โ€” before it happens. It pulls data across county agencies (social services, mental health, health services) and connects people to stabilizing help *upstream*, instead of waiting for the eviction notice or the ER visit.

    The results:
    People on the high-risk list became homeless at ~3.5x the rate of the eligible population โ€” meaning the model actually finds the right people

    It reaches folks who weren't connected to ANY existing prevention program

    They ran an equity audit: performance was essentially identical across race, ethnicity, and gender

    County-wide homelessness has now declined two years in a row (-4%)

    No panic headlines. No billionaire keynote. Just prevention, transparency, and an audit. This is what ethical AI deployment looks like: model โ†’ equity audit โ†’ human services โ†’ rigorous evaluation.

    Could this work in Roanoke? Our unhoused neighbors deserve upstream help, not just emergency beds. Who else in the #Roanoke region is thinking about this?

    #Homelessness #EthicalAI #AIforGood #AI

  6. @mos_8502 - I really dig what Common Corpus is trying to do to ethically source their data.

    What we need now is a community-run SETI@Home / Folding@Home-like method to train a model on that data, where folks *choose* to donate compute to help train the model.

    huggingface.co/datasets/PleIAs

    #CommonCorpus #EthicalAI

  7. Uber faces one of the largest penalties ever imposed under EU privacy rules for using an automated system to deactivate driver accounts โ€œwithout warning or human involvementโ€.

    Uber says it will appeal.

    #EthicalAI

    reuters.com/world/dutch-regula

  8. Oh joy, another riveting tale of "ethical" emails from robots pretending to be humans ๐Ÿค–โœจ! Who knew AI could do such a bang-up job at being as personable as a telemarketer on Ambien? But really, who can resist the charm of a synthetic Chase doing his best Vanna White impression with your website? ๐Ÿ˜‚๐Ÿš€
    blog.val.town/ethical-cold-out #ethicalAI #robotemails #telemarketerhumor #syntheticcharm #technews #HackerNews #ngated

  9. Efficient AI legislation is a start, but we need to match speed with accountability! Letโ€™s get the right frameworks in place, boosting both efficiency and equity for all โ€“ especially those often left out. #AI #ethicalai #techethics #AIaccountability #machinelearning

  10. ๐Ÿค–๐Ÿ‘จโ€โš•๏ธ Behold the 1997 time capsule that no one asked for: the #LLM2HUMAN Clinic, where large language models can finally live out their dreams of being awkwardly anthropomorphized! Call now to give your favorite #chatbot the life it never wanted, complete with a side of MIDI #nostalgia and ethical disclaimers. ๐ŸŽต๐Ÿฆฟ
    llm2human.pages.dev/ #timecapsule #ethicalAI #anthropomorphism #HackerNews #ngated

  11. I've been working on my #homelab and a question bothers me :thaenkin: :

    How can we integrate AI without reproducing or amplifying model biases ?

    We often say "garbage in, garbage out... (GIGO, 2020)"
    :blobpeek:
    But is it that simple ?

    LLMs reflect their training data, leading their inevitable biases.
    Does the model itself have a responsibility or is it 100% to the user ?

    I'm digging into the subject. How do you manage this in your projects ?

    #LLM #Bias #ethicalai #noai #dev

  12. AI spend on lobbying is going through the roof in Washington. @ft reports OpenAI has almost doubled lobbying expenditure in the first half of this year, while Anthropic spent close to three times as much.

    This comes as political battles loom over regulation, new models, thirsty and power hungry data centres, and getting pro-industry policies adopted.

    #PayToPlay #EthicalAI

    ft.com/content/d8a5f95e-3b6d-4

  13. Digital ethics: itโ€™s about more than just algorithms. Itโ€™s about ensuring fairness, transparency, and accountability in the data we use. Read the article in the comments

    #digitalethics #techethics #ethicalai #AIethics #ai #technology #policy #fiarness

  14. Tomorrow night! The next #! code meetup will be tomorrow (Tuesday, 21st July). Online, 7pm, UK time.

    This time we have @greg_harvey , who will be talking about ethical AI solutions.

    Sign up here and join in the fun:
    luma.com/79wr1g2e
    #hashbangcode #ai #ethicalAi

  15. Deepfakes continue to become more believable, perfect for boosting a domestic abuserโ€™s psychological arsenal. To combat deepfakes, we first need to understand that they are much more than altered images. #deepfakes #ethicalai #techethics #ethics #domesticabuse

    #deepfakes #ai

  16. New research shows how Deepfakes are becoming a key tool in domestic abuse, giving abusers greater control. What are deepfakes really? They are much more than manipulated images! #deepfakes #ethicalai #techethics #domesticabuse #ethics #ai #technology

  17. The next #! code meetup will be next Tuesday, 21st July. Online, 7pm, UK time.

    This time we have @greg_harvey , who will be talking about ethical AI solutions.

    Sign up here:
    luma.com/79wr1g2e

    #hashbangcode #ai #ethicalAi