Nam Le

@namlehai

Cognitive Science&AI geek. Musician. Interested in Evolutionary & Neural Computation, Artificial Life, Cultural Evolution, Evolutionary Psychology&Epistemology

Dublin City, Ireland
Vrijeme pridruživanja: svibanj 2012.
60 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @namlehai

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

  1. Prikvačeni tweet
    22. kol 2018.

    Agents (blue-purple) have to detect and classify food (red) and toxin (white), affecting the direction of the agent (Braitenberg love and hate). without reinforcement signals, Hebbian is the only way to do. Hebbian+Evo population can detect and eat food, while evo alone fails.

    5 proslijeđenih tweetova 20 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  2. 24. sij

    Artificial Life generating arts

    Lots of people today talk about using artificial intelligence as a creative tool in the arts, but artifical life can do amazing things too. (I don't know exactly how generates these beautiful movies, but I assume it's using a-life techniques.)
    1 korisnik označava da mu se sviđa
    Poništi
  3. proslijedio/la je Tweet
    6. sij

    I am extremely excited to announce (1) I've joined OpenAI to lead a large-scale effort into AI-generating Algorithms research, & (2) I'll be an Associate CS Professor at U. British Columbia in 2021, where I will continue to lead the OpenAI project. Both are dreams come true! 1/2

    89 replies 120 proslijeđenih tweetova 2.085 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    1. sij

    Just watched this AI DEBATE with Yoshua Bengio and and was inspired to write a blogpost about innate priors and how much information is encoded in the genome (coming soon...)

    17 proslijeđenih tweetova 117 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. 15. pro 2019.

    Meta-learning by Evolution, assimilation and accommodation of behaviour, the Baldwin Effect through gene-culture coevolution is the key. Cultural information is partly encoded as some degree of predisposition to learn it.

    A 4 year old child actually has a few hundred million years of experience, not 4. Their rapid learning/generalization is much less shocking/magical considering this fact.
    Prikaži ovu nit
    1 proslijeđeni tweet 5 korisnika označava da im se sviđa
    Poništi
  6. 29. stu 2019.

    Thinking about the difference between education and training, I see another reason why the current machine learning/AI by training cannot reach the intelligence level of an educated being.

    1 reply 3 korisnika označavaju da im se sviđa
    Poništi
  7. 24. stu 2019.

    Today, 24 November 2019, is Evolution’s Day — 160 years since the first publication of The Origin of Species by Charles Darwin. This landmark changed and still continues to affect the way we think about ourselves and about the world we live.

    1 proslijeđeni tweet 1 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  8. 16. stu 2019.

    Bias-variance dilemma phrased by The Cognitive Scientist Peter Todd: The human cognitive system can make good generalization despite sparse samples because the variance component is controlled within a limit => not general-purpose inductive inference, but specialization.

    4 korisnika označavaju da im se sviđa
    Poništi
  9. 11. stu 2019.

    In a Classical Guitar Master Class, a book called “Playing with ease” written by the master class teacher. Interesting to see how bodily movements change and affect music creativity. Embodiment is important even for something unique to human intelligence, creativity.

    1 korisnik označava da mu se sviđa
    Poništi
  10. proslijedio/la je Tweet
    9. stu 2019.

    What is the least amount of innate machinery we would need to build an intelligent creature? My proposal from 2018 [] is below. (Convolution=translational invariance) cc

    34 proslijeđena tweeta 88 korisnika označava da im se sviđa
    Poništi
  11. 10. stu 2019.

    Innateness is necessary, and of course created by Evolution. I just want to add: "moving through a high dimensional fitness landscape without a gradient is really inefficient" -- We have evolved our learning ability to cope with this. And the Baldwin Effect presents how it works.

    Yes! Innate machinery is achieved through natural selection. The challenge for simulating evolution is that moving through a high dimensional fitness landscape without a gradient is really inefficient. Evolution has solved this by brute force--at least 10^23 animals have existed
    1 reply 1 proslijeđeni tweet 5 korisnika označava da im se sviđa
    Poništi
  12. 5. stu 2019.

    Adam Smith — an Enlightenment thinker: “Poverty needs no explanation. What needs to be explain is the wealth”. In the world governed by evolution and entropy, Humans exploit energy to resist entropy, maintaining their own equilibria, releasing messes into the world.

    1 reply 3 korisnika označavaju da im se sviđa
    Poništi
  13. 22. lis 2019.

    A funny metaphor. The way most Machine Learning adopt today is pretty similar to the Confucius philosophy of learning, dominant in Eastern Asian, with tons of examples from the teacher, with fewer creativity. This is, of course, different from Enlightenment philosophy of learning

    2 korisnika označavaju da im se sviđa
    Poništi
  14. 4. lis 2019.

    Most foundational advances of today came from Cognitive Science guys, and most current AI critics are Cognitive Scientists. There is a huge gap between engineering side (mostly endowed with Machine Learning and CS background) and foundational path towards intelligence.

    4 korisnika označavaju da im se sviđa
    Poništi
  15. 28. ruj 2019.

    Reading “Rebooting AI: Building Artificial Intelligence We Can Trust” by . Gary has spoken out in recent years about the misunderstanding and over-optimism of the public for the deep learning movement. Will give feedbacks. Mind the Gap reminds me of the London Tube.

    1 reply 1 proslijeđeni tweet 3 korisnika označavaju da im se sviđa
    Poništi
  16. 27. ruj 2019.

    Is "factor g" for general intelligence a myth? Has general intelligence ever existed? If never, AGI should change its approach. Human intelligence seems not to be a plausible model.

    2 korisnika označavaju da im se sviđa
    Poništi
  17. 1. kol 2019.

    A book to be read. Yet I hope I can find something new and different from what has been discussed by many in recent years about the role of phenotypic plasticity and development in evolution and the reciprocal causality between them.

    Our book "Evolutionary Causation: Biological and Philosophical Reflections" (MIT) will come out in Sept! We'll be tweeting about the chapters when the publisher links go live. Until then, here's a sneak peek of the contents:
    3 korisnika označavaju da im se sviđa
    Poništi
  18. 28. srp 2019.

    Evolving AI. My point here is that Intelligence is not merely a passively evolved trait of man and certain other animals, but is itself a process of evolution. Intelligence should be seen as both cause and consequence of evolution. This is an open-ended process.

    3 proslijeđena tweeta 12 korisnika označava da im se sviđa
    Poništi
  19. proslijedio/la je Tweet
    26. srp 2019.

    A "worrying analysis": "18 [] algorithms ... presented at top-level research conferences ... Only 7 of them could be reproduced w/ reasonable effort ... 6 of them can often be outperformed w/ comparably simple heuristic methods." Paper:

    1.283 proslijeđena tweeta 2.597 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  20. 23. srp 2019.

    Just read this fun book on the flight from Budapest back to Dublin. Acient Greek myths seem to be a great resources for AI and ALIFE studies.

    1 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  21. 23. srp 2019.

    1 reply 2 proslijeđena tweeta 5 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi

@namlehai još nije objavio nijedan Tweet.

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:

    ·