hdarshane

6 ms·

1.
Thinking out loud: evolution and pretraining(hiranmay.com)
2 points by hdarshane 7mo ago|0 comments
2.
Physics could benefit borrowing a little from AlexNet and Sutton's Bitter Lesson(hiranmay.xyz)
1 points by hdarshane 2y ago|1 comments
3.
Physics could benefit borrowing a little from AlexNet and Sutton's Bitter Lesson(hiranmay.xyz)
2 points by hdarshane 2y ago|0 comments
4.
The Techno-Humanist Manifesto: The Surrender of the Gods, Part 1(blog.rootsofprogress.org)
1 points by hdarshane 2y ago|0 comments
5.
HCI Is the Bottleneck(joshmpollock.com)
2 points by hdarshane 2y ago|1 comments
6.
by hdarshane 2y ago
Thank you so much for the link to those two papers. I'll try and go through them. >Like, they undergo many cell cycles during one organism's life, but they're not really evolving or being selected during each cell cycle. T
7.
by hdarshane 2y ago
>This is especially tricky in the field of bio / biochem, which feels like it requires memorizing a million facts before one gets to do any real thinking and reasoning. This unfortunately tends to filter out a lot of more mathematic
8.
by hdarshane 2y ago
Huge thanks for the list! Interesting to wonder just how influential or generally just of what nature cybernetics would be today without funding issues, lack of categorisability and the field slowly drifting away from the original vision Wi
9.
by hdarshane 2y ago
Thanks a lot for your kind words! Really like your thoughts! Indeed, lack of time-series observability makes it harder for us to find general patterns or causal events. Definitely agree that biology is the pinnacle of all complexity - IMO s
10.
by hdarshane 2y ago
Thanks a lot! I think that tool might be just any general framework that gives a better idea of what any small microscopic actions might lead to at a high-level. We definitely cannot qua(nt/l)ify how actions seep through given free-wil
11.
by hdarshane 2y ago
Thank you so much for the kind words! I'm definitely a little ignorant with respect to systems biology efforts. I have an okay-ish idea of the work of someone like Uri Alon, it's definitely not enough. What resource would you reco
12.
by hdarshane 2y ago
Oof yeah. The legibility-gained per effort put in for most admins working in a system that inherently incentivises tangibility and observable "utility" (whatever that may be in this case) reduces any hope of seeing much change. Ma
13.
by hdarshane 2y ago
Yeah, entropy definitely has been included in some forms of thought that are very correlated with complex systems. Schrodinger early on had a very interesting insight on life and entropy, people like Jeremy England are taking that view forw
14.
by hdarshane 2y ago
Wow. This is a really informative comment. >While I agree that nitty gritty formalizations are important to grasp, it's been made clear that the formal sciences were often insufficient in facilitating insight on the type of complexi
15.
by hdarshane 2y ago
Thanks a lot! SO many interesting recommendations in your comment. Yes, I agree that divergence is a very positive property right now, it might also allow for even more unexpected cross-pollinations and comparative studies as the pool of st
16.
by hdarshane 2y ago
Interesting! Have heard about Christopher Alexander's work, but have never really jumped in. Maybe I should do that now. >I'm of the opinion right now that what we call "design" and "architecture" is really
17.
by hdarshane 2y ago
Yes. I'm aware. I think cybernetics has definitely drifted away from what the original scope was, also I think modern interpretations of cybernetics-ish ideas and complex system studies are slowly also incorporating social sciences an
18.
by hdarshane 2y ago
Yes! I think it was a parallel study between nervous systems and electrical engineering principles by Wiener et al. that kickstarted cybernetics? Interesting! Will ping you!
19.
by hdarshane 2y ago
Absolutely. Broadly, I wonder what leads to schools not adopting emerging fields as part of the formal curriculum. Interesting to note even in 2021, only 51% US k12 high schools were found to have a CS course. Does not seem like a capital
20.
by hdarshane 2y ago
Thank you for your very interesting comment and the reading recommendation. I agree with your general sentiment that chasing "wide applicability" or trying to force a narrative that xyz theory will explain xyz might be hugely detr
21.
We must seek a widely-applicable science of systems(hiranmay.xyz)
81 points by hdarshane 2y ago|67 comments
22.
AI Agents for Desktop Computing(hiranmay.xyz)
2 points by hdarshane 2y ago|0 comments
23.
by hdarshane 2y ago
Non-deterministic nature of these models is definitely a concern. Interesting to see how companies implementing generative interface circumvent this. Maybe they create a set of fixed generic UI elements or write a system prompt describing g
24.
by hdarshane 2y ago
Check out some of the work at https://www.inkandswitch.com/patchwork/notebook/
25.
WebSim: Hallucinate an Alternate Internet with Claude 3 [video](youtube.com)
1 points by hdarshane 2y ago|0 comments
26.
by hdarshane 2y ago
I think the importance of this study is being downplayed by many. That same spontaneous and "breakthrough" nature of the abilities of LLM that is now nullified, was a major component of many arguments against LLMs and could'v
27.
by hdarshane 2y ago
This interpretation is most definitely right. As with all things, it's coming down to semantics.
28.
by hdarshane 2y ago
emergence simply means something the model wasn't trained and expected to produce.
29.
by hdarshane 3y ago
I'm not Alan, but I'm pretty sure he isn't too happy about it.
30.
by hdarshane 3y ago
>A first-class programming language (not an LLM) to talk to the computer's OS along with a rich library is the most important missing component IMO. That "programming language" might just be ability to have generative inte

More ›