Alex_Notchenko

Joined 14 years ago

Karma

Submissions

Member since

Aug 23

Previously: Head of AI at brask.ai PhD in 3D deep learning for scene reconstruction

avnotchenko [circled 'a' letter] google mail domain [end of sentence symbol] COMercial top level domain

View on Hacker News

Recent submissions

Comment · 17 days ago

Keep in mind if Alexandra Elbakyan shares the knowledge she is a criminal who needs to be extradited to face consequences. When you are a Californian AI lab that needs to train a model to give back all this static information to users, you are alright, actually you are the only thing keeping US economy going. Never be in out group its hard out there.

Comment · 10 months ago

Sqlite database on your local macbook is encrypted so there is no way to export data as far as I'm aware, I want to get access to it, and migrate to screenpipe. If anyone knows how to do an export of local database, please let me know.

Comment · 1 year ago

if you want to learn more about modern luddites check out "This Machine Kills" podcast and to some extent Ed Zitron / Cory Doctorow Blogs, might be a good place to start.

Political-Economic analysis of technology is not super popular thing in a mainstream media, but disabling, sabotaging or vandalising anti-human tech might be.

Comment · 1 year ago

I run local communities (specifically a London based Russian Language Machine Learning and AI community, hit me up if you are interested) it's 100% cannot survive long term without a locus in physical space. Online only exists to facilitate offline, if people are just avatars and not real people, there is a hard limit to what can be achieved in a community, its a question of trust.

Comment · 4 years ago

In modern deep learning there are multiple approaches that can be argued are close to some of inspirations from neuroscience: Capsule Networks, Helmholtz Machines, Energy Based Models (Score Based Generative Modeling / Diffusion Models), Associative Memory.

Information theory, Bayesian methods, Approximate Computations are more relevant for inspiration. Neuroscience is not the filed which studies intelligent behavior.

Comment · 11 years ago

You right. For instance, before huge drop in value of Ruble, people rushed to buy all kinds of goods (Big TV's, Apple products, Cars) to capture value of currency. It can be argued that sudden increase in purchases can signal anticipation of economic downturn.

Comment · 11 years ago

That is, when it's not used for echolocation. When we arrive at more scientific definition of "neuroplasticity" (or whatever the underlying effect is called) this will be the one of examples.

https://en.wikipedia.org/wiki/Human_echolocation

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3102086/

PS: There's also great episode of Radiolab on that topic.

http://www.npr.org/sections/13.7/2013/01/28/170355712/be-lik...

Comment · 11 years ago

Article doesn't have "history" in it's title, HN submission has. Title on site is "How the Computer Got Its Revenge on the Soviet Union".

> some cheap spin-doctoring on soviet computer science this article is actually reworked version of "Gerovitch, Slava. "InterNyet: why the Soviet Union did not build a nationwide computer network." History and Technology 24.4 (2008): 335-350." which has a lot of reference and citations. And article main subject is not history of computer science in USSR, it's about problems of creating huge computer networks in particular political-ideological environment.

I don't see how this attempt of informing about attempts of creating Internet in USSR diminishes achievements of great Scientists that you've mentioned. It mentions "MESM", which was created by Lebedev, it's certainly close enough without changing the subject of article.

> Twisting the facts, mixing them with fiction in such a way that is hard to tell one from another. Where exactly fact twisting is in the article?

[0] http://web.mit.edu/slava/homepage/articles/Gerovitch-InterNy...

Comment · 11 years ago

There are several ideas on how to simulate social network experience with classical w3c tech. Linkeddata group (one with Tim Berners-Lee) work on SoLiD - "for developers who plan to build social linked data servers and applications". As you said, there is plenty of ways to make online social experience with original WWW methods, and it'll be distributed, and ad free, you just need to make a "pretty wrapper" for common folks.

If more developers acknowledged possibilities of semantic web (or "Linked Data" approach as it been re-branded) not just for reducing entropy of the web, but also to provide solid services for ordinary users, it wouldn't be so hard to develop critical mass to be widely used.