Sean McClure

@sean_a_mcclure

Founder Kedion, Ph.D. Quantum Chem, builds AI software, studies real-world complexity.

San Francisco, California
Vrijeme pridruživanja: siječanj 2011.

Tweetovi

Blokirali ste korisnika/cu @sean_a_mcclure

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

  1. Prikvačeni tweet
    28. srp 2019.
    Odgovor korisnicima i sljedećem broju korisnika:

    Reality is nonlinear, involving higher dimensions than anything simplistic statistical techniques can model correctly. If you want to model reality you need to use high-dim approaches that can adequately approximate complexity.

    Poništi
  2. proslijedio/la je Tweet
    prije 15 sati

    Today’s . wealth -> intelligence -> wealth -> intelligence -> ... umh, wha? | inspired by For more visit For my books

    Poništi
  3. proslijedio/la je Tweet

    Great artists don’t need awards.

    Poništi
  4. 31. sij

    Thank you. Always great to hear back from those positively-impacted. Twitter needs more of this. Keep embracing the self-education 👍 True skills only emerge from your efforts to build something tangible.

    Poništi
  5. 31. sij

    The source of hype in AI is our projection of narratives onto its outputs. If AI writes a poem we find meaning. If it paints we see art. It makes sound we hear music. It’s not so much the narrow limitations of AI that are problematic, it’s our inescapable attraction to story.

    Poništi
  6. 30. sij

    Why is Slower Faster? Because working quickly makes you spend a ton of energy scratching at the surface, never uncovering the few deep realizations needed to repeatedly turn raw input into incredible output. The usable pieces are deep. You will not find mastery on the surface.

    Poništi
  7. 30. sij
    Prikaži ovu nit
    Poništi
  8. 30. sij

    Congratulations to Mr. Pinker for becoming immortalized by the likes of Taleb. Well deserved.

    Prikaži ovu nit
    Poništi
  9. 30. sij

    However, stating that techniques like MDA do not rely on strong assumptions is an overstatement. Knowing that a variable plays a role is not the same as knowing what role. The decision to be made in model interpretation is to keep/remove. Much more cmplx than knwng it plys a role

    Prikaži ovu nit
    Poništi
  10. 30. sij

    highlights 2 major points: 1. difference between in-sample vs out-of-sample validation; 2. the reuse of datasets by social scientists. Note, social scientists don't get that in-sample fitting is NOT prediction. Also, they don't understand data leakage.

    Prikaži ovu nit
    Poništi
  11. 29. sij

    Don’t brag about being well read. Reading requires almost no cognitive effort. Wanting to be respected for the mindless consumption of someone else’s effort is embarrassing. When you build stuff you don’t need to brag, you just point.

    Poništi
  12. 29. sij

    The community has spoken. Most important topics to know, mapped out with Wikis for each topic. Read and learn.

    Poništi
  13. 27. sij

    Emotion is highly effective as a “lubricant of reason” but there needs to be something to lubricate. Jumping immediately into angry statements backed by nothing only garners support from the cognitively lazy. You still have to start from reason to go beyond its limitations.

    Poništi
  14. 27. sij

    There are only so many distinct concepts. This is why going deep into any one field can uncover universal truths that span seemingly distinct domains.

    Poništi
  15. 27. sij

    In other words, it’s perfectly fine to believe human cognition is something more than “machine” but you have to start your argument with an understanding of what a machine is capable of. If you artificially stymie the definition of machine then your argument is based off fls prms

    Poništi
  16. 27. sij

    This is on a good day. Most don’t make an argument period.

    Poništi
  17. 27. sij

    Great feedback from ...thanks again for asking me to weigh-in. Looking forward to continued discussions.

    Poništi
  18. 23. sij

    Decadence has forced us to artificially recreate what nature gave us for free.

    Poništi
  19. 23. sij

    Exercise is important. But there’s something sad about forcing yourself to simulate stress using fashioned weights at a gym. Zoos recreate environments by providing behavioral opportunities. That’s your gym. A sterile room designed with “behavioral opportunities.” Sad.

    Poništi
  20. 23. sij

    Be wary of calm perspectives. They are usually borrowed, by people who haven't lived the life necessary to warrant their point of view. Spending your life perched on a remote mountain calmly contemplating existence only means something if you climbed the mountain.

    Poništi
  21. 23. sij

    When you overhear people at the coffee shop spending their entire conversation complaining about people, you can be fairly certain they are accomplishing nothing with their lives.

    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:

    ·