Emanuil

@emanuil_tolev

Empathy. . Ex small agency lead . Open data & science. . Newbie runner & archer. He/him.

London / Cardiff
Vrijeme pridruživanja: svibanj 2010.

Tweetovi

Blokirali ste korisnika/cu @emanuil_tolev

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @emanuil_tolev

  1. prije 12 sati

    Do we need an ethical API? (for diff kind of ethics and to define how ethics interact with machines) A final reflective q for !

    Prikaži ovu nit
    Poništi
  2. prije 12 sati

    Q: Are governments capable of regulating this market? They don’t understand tech on a very basic level. UK govt is pushing business to do more rather than more regulation and that seems Ok at the moment but it’s very difficult

    Prikaži ovu nit
    Poništi
  3. prije 12 sati

    Q: pessimistically assume that companies will be evil. Now what? Perhaps we agree on all of us not using certain techniques like face detection for ex, and regulation.

    Prikaži ovu nit
    Poništi
  4. prije 12 sati

    “Don’t be evil” - how do orgs promote a culture of actively doing good? Appreciating social cohesion (volunteering, holidays, etc). Having people accountable in orgs for aspects of AI. Setting clear principles and taking them seriously.

    Prikaži ovu nit
    Poništi
  5. prije 12 sati

    Heh heh, first q “how do you ensure an open source tool is used responsibly”. It’s difficult to do so and many tools are already OSS.

    Prikaži ovu nit
    Poništi
  6. prije 12 sati

    Can’t just sit back and say “well I’m just a dev/designer/etc” - that isn’t good enough anymore and never was, we have to take responsibility.

    Prikaži ovu nit
    Poništi
  7. prije 12 sati

    - create an AI governance group: designers, users, directors, data scientists - risk assess your service against ethical issues and unintended consequences - revisit & update your AI processes with new tech - publicly declare your intents, processes and updates

    Prikaži ovu nit
    Poništi
  8. prije 12 sati

    But it’s not good enough to publish your principles — Project IF. Businesses need to work on services in society’s interest - rethink what you’re doing otherwise. In practice:

    Prikaži ovu nit
    Poništi
  9. prije 12 sati

    Multiple people at Amazon want to intro these topics but find it somewhat difficult due to the org culture. OECD guidelines on AI particularly good: benefit people and planet, respect humans, safeguards, be transparent, robust and accountable.

    Prikaži ovu nit
    Poništi
  10. prije 12 sati

    Do we need an ethical API? (for diff kind of ethics and to define how ethics interact with machines) A final reflective q for

    Prikaži ovu nit
    Poništi
  11. prije 12 sati

    Ethics must become the centrepiece of any organisation developing AI. We have to reflect on what we’re doing or we walk blindly into a future that we do control. (heh!) Tech ethics, data ethics and machine ethics (how we code ethics into any machine) are all relevant.

    Prikaži ovu nit
    Poništi
  12. prije 12 sati

    “Facebook manipulated 100’000s of users’ emotions for science” headline (gonna check that one later).

    Prikaži ovu nit
    Poništi
  13. prije 13 sati

    Anybody got info or example rubrics for SoC (security operations centre) analyst interviews? Helping jr techie prep. If you've done that job and would like to chat to her to help her a little to get into the field also lmk.

    Poništi
  14. prije 18 sati

    As an aside, if you read this thread from the PoV of a woman telling a random guy on here he needs therapy, it will read as policing how she can express her frustration. I don’t like or want to do that but even then the core idea is true to me 🤔.

    Prikaži ovu nit
    Poništi
  15. prije 18 sati

    You most often end up implying there’s sth intrinsically wrong with them and that’s why they <did a stupid thing / hurt someone / can never right their wrongs>. This is insidious and works against everyone doing better. Just reconsider how you use “you need therapy”.

    Prikaži ovu nit
    Poništi
  16. prije 18 sati

    Of course you don’t have to be useful to others all the time, you’re as entitled to complaining as anyone else if others are annoying. But using *therapy* as the edge of that crystallised frustration has bad effects you’re probably not considering in the moment of complaining.

    Prikaži ovu nit
    Poništi
  17. prije 18 sati

    Explaining (seriously, jokingly, metaphorically - whatever) that you think the person has made a mistake? Their actions don’t align with what you think their values are? Don’t align with who you want them to be? All useful. Just going “you’re broken and invalid!!” - not so much.

    Prikaži ovu nit
    Poništi
  18. prije 18 sati

    The more I do therapy the weirder I find its weaponisation on social media. Shouting “You need therapy!!” when you’re angry at someone doing something that’s stupid to you is .. the opposite of what therapy would teach you to do. Do you also shout at yourself? It doesn’t help.

    Prikaži ovu nit
    Poništi
  19. 2. velj

    Really interesting to hang out with some of the crew of . They’re holding their first FOSS conference in 2 weeks. Very rejuvenating to talk to people planning their first event as we conclude our 20th and my 4th as an organiser. I’m in a very contemplative mood.

    Poništi
  20. proslijedio/la je Tweet
    1. velj

    People really struggle to separate the concepts of "you have an obligation to do X", "it is morally virtuous to do X", "it is practically beneficial to do X", and "you must do X". In fact none of these imply any of the others

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·