andreyk

39 ms·

421.
by andreyk 5y ago
I don't believe so, as it's more about competitive games and about incomplete information. They did release Agent 57 last year, and that got better than human results on all Atari games (though still not sample efficiently). More
422.
by andreyk 5y ago
Here's a link to the original paper for those who might prefer it: https://link.springer.com/article/10.1007/s10458-021-09514-w This is a slightly edited and updated version of it.
423.
How to train your decision-making AIs(thegradient.pub)
54 points by andreyk 5y ago|5 comments
424.
Rishi Bommasani on Foundation Models(thegradientpub.substack.com)
1 points by andreyk 5y ago|0 comments
425.
by andreyk 5y ago
FYI, there are a ton of definitions of 'intelligence', some fit with what AI algorithms can do, some don't. I found reading 'A Collection of Definitions of Intelligence' quite interesting myself https://a
426.
by andreyk 5y ago
I actually tried to do this with Diaspora and a couple dozen of my closest friends. We tweaked it to not have like buttons, and like other fun stuff. The idea being that it'd be our own little space and we could be weird on there and d
427.
by andreyk 5y ago
IMHO (as a AI researcher) 'multi-modality, multi-task, multi-language, multi-memory' may take longer than you think. Doing multi modal learning is quite hard due to lack of labeled data, a lot of tasks either don't offer clea
428.
Bots, Lies, and DeepFakes – Online Misinformation and AI's Role in It(skynettoday.com)
1 points by andreyk 5y ago|0 comments
429.
by andreyk 5y ago
Ugh the title of the article. Here's the paper that this article is recapping - "Reduced, Reused and Recycled: The Life of a Dataset in Machine Learning Research", https://openreview.net/forum?id=zNQBIBKJRkd
430.
Upol Ehsan on Human-Centered Explainable AI and Social Transparency(thegradient.pub)
1 points by andreyk 5y ago|0 comments
431.
Upol Ehsan on Human-Centered Explainable AI and Social Transparency(thegradient.pub)
1 points by andreyk 5y ago|0 comments
432.
by andreyk 5y ago
Fun fact, there is a podcast with this concept (but not AI based) - Boring Books for Bedtime. I love it! Use it to help me sleep all the time.
433.
Upol Ehsan on Human-Centered Explainable AI and Social Transparency(thegradient.pub)
1 points by andreyk 5y ago|0 comments
434.
by andreyk 5y ago
Hmm, I've been a user of this kind of thing for a long time and there's quite a few 'habit-tracking' apps out there (my personal choice being Loop Habit Tracker, I dig its minimalism). Is your key differentiator the acco
435.
Things Everyone Should Know About Depression(andreykurenkov.com)
3 points by andreyk 5y ago|0 comments
436.
by andreyk 5y ago
Good list! As a PhD student and therefore AI researcher for a few years now, a lot of this rings true. Though 100 lessons is too much and some of these are obvious/minor, i'd distill it down to the main ones. Here's my 2 cent
437.
by andreyk 5y ago
This is a common problem with language models in general, and the reason is not that well understood. This paper A Theoretical Analysis of the Repetition Problem in Text Generation ( https://arxiv.org/abs/2012.14660 ) se
438.
by andreyk 5y ago
Why would I write an article about my own limitations :P
439.
by andreyk 5y ago
Good point, thanks!
440.
by andreyk 5y ago
That's fair , I forgot how many they had vs just saying it is powering 300 apps. There is also http://gpt3demos.com/ with lots of live demos and varied things, though it's more noisy.
441.
by andreyk 5y ago
OpenAI has a blog post highlighting many (edit, not many, just a few) applications - https://openai.com/blog/gpt-3-apps/ It's quite powerful and has many cool uses IMHO.
442.
by andreyk 5y ago
yep, pretty much what i'm saying here. Though not all language models are built the same, eg the inference cost is unique to it due to its size. Still, most of this applies to any typical language model.
443.
by andreyk 5y ago
Author here, would love feedback / thoughts / corrections!
444.
by andreyk 5y ago
This is very specific to GPT-3 and not generally true though. And GPT-3 is not an agent per se but rather a passive model (it received input and produces output, and does not continuously interact with its environment). So it makes sense in
445.
The Inherent Limitations of GPT-3(lastweekin.ai)
88 points by andreyk 5y ago|123 comments
446.
The Inherent Limitations of GPT-3(lastweekin.ai)
3 points by andreyk 5y ago|0 comments
447.
The Inherent Limitations of GPT-3(lastweekin.ai)
3 points by andreyk 5y ago|1 comments
448.
OpenAI's Miles Brundage on AI Misuse and Trustworthy AI(thegradientpub.substack.com)
4 points by andreyk 5y ago|0 comments
449.
by andreyk 5y ago
I'd guess a lot of it is cultural - I believe much of Europe also has a far less workaholic culture than the US, with more emphasis on living life and whatnot. And I'm sure there are other factors (France's labor laws presuma
450.
Job Loss Due to AI – How Bad Is It Going to Be?(skynettoday.com)
1 points by andreyk 5y ago|1 comments

More ›