Andrew Trask

@iamtrask

Helping people answer questions using data they cannot see Lead , SRS , PhD S. +, , Edu +

Oxford, UK
Joined November 2012

Tweets

You blocked @iamtrask

Are you sure you want to view these Tweets? Viewing Tweets won't unblock @iamtrask

  1. Retweeted
    Jan 22

    OpenMined + collaborating to advance open source software development. Learn about these talented teams on our blog: Many opportunities ahead. Join our Slack community to find out more!

    Undo
  2. Retweeted
    Jan 30

    Welcome OpenAI to the PyTorch community!

    Undo
  3. Retweeted
    Jan 19

    Here's a lecture by Andrew Trask () on privacy-preserving AI as part of the MIT Deep Learning lecture series. Preserving privacy boosts our ability to do science at a large-scale and to engineer intelligent systems that learn from data:

    Undo
  4. Jan 28

    IMO, this is the best work I've ever read on AI & Human Values Highly recommended read for everyone in the field It's a dense read - but push through. You'll be glad you did.

    Undo
  5. Retweeted
    Jan 26

    If you want to learn about privacy-preserving machine learning, then there is no better resource than this step-by-step notebook tutorial by . From the basics of private deep learning to building secure ML classifiers using PyTorch & PySyft.

    Undo
  6. Jan 26

    Commercial Encyclopaedia -> Commercial Software -> Commercial AI -> ???? Intelligence has a history of becoming free on the internet. AI will be no different.

    Undo
  7. Jan 26

    In the future - you will be able to own and control the only copies of your information While still receiving the same goods and services you receive now. Let's build the future we want to live in Don't know where to start? Join

    Undo
  8. Jan 26
    Undo
  9. Jan 25

    When I learn new idea, I must repeatedly learn it from multiple before it *clicks* If you're learning nets - add this to your list It discusses for: - images - audio - databases

    Undo
  10. Jan 24

    is one of the most important techniques I don't often recommend PhD Thesis' - 's is exceptional. He's a brilliant writer! Check out this taxonomy / table of contents!!! 👇👇👇

    Undo
  11. Jan 23

    My phone should check each image I take for skin The model should be: - Free - Apache 2 licensed - Crowdsourced using Similar to what did for software This is what wants to do for AI Join us!

    Undo
  12. Jan 23

    If *anyone* would be un-hackable, surely it would be: 1) the world's wealthiest person 2) who leads one of the world's largest and most secure cloud platforms 3) running the world's second most popular phone OS Lesson learned: no-one is un-hackable.

    Undo
  13. Jan 22

    Googling Tip: when you want to learn some ML concept, google "python <concept name> from scratch". This will often lead to a blogpost with: - intuitive explanation - a toy implementation - reference to more formal papers which you can unpack in that order :)

    Undo
  14. Retweeted
    Jan 18

    Today we typically use centralized learning, but: - Centralized learning is good when you can bring the data to the model; - Federated learning is good when you can bring the model to the data; - Encrypted learning is for when you can’t do either.

    Undo
  15. Jan 18

    These guys scraped social media and created a smartphone app for facial recognition. "Sure, that might lead to a dystopian future or something, but you can’t ban it" "Clearview’s investors predict that its app will eventually be available to the public"

    Undo
  16. Jan 18

    Papers are published online - conferences should be too!!! More attendees ($30 instead of $2000+, no visa issues) More content (anyone could stream a talk) More often ( every quarter) Seems like a no brainer Would you go to an ?

    Undo
  17. Jan 17

    Many experts have heard of , which is data-parallel Equally compelling, but less well known is , which is model-parallel Here gives a step-by-step implementing ! A great piece!

    Undo
  18. Jan 17

    After a year of dev, I am *extremely* excited to share this step-by-step tutorial Goal: to be the *easiest* intro to preserving, Deep Learning It's in I hope you enjoy it

    Undo
  19. Jan 16

    If you've wondered - "Which optimizer should I use? ? ? ?" This blogpost by is the best explanation I've seen. It's a surprisingly easy read! Definitely a great / project!

    Undo
  20. Jan 15

    The *easiest* way to learn Deep Learning is to build it from scratch! IMO, the same is true when learning a Deep Learning framework. In I show how to build a -like framework Here's the step-by-step code!!

    Undo

Loading seems to be taking a while.

Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

    You may also like

    ·