atomashevic

Linux option for Codex App https://github.com/ilysenko/codex-desktop-linux

Mobile remote connection works, pushed the PR earlier today.

Yes it does. Codex, three dots menu: Codex cloud

    Location: Europe, Serbia
    Remote: Yes
    Willing to relocate: No
    Technologies: Python, Transformers, TensorFlow, Keras, pandas, scikit-learn, R, Linux, HTML/CSS, Git
    Résumé/CV: https://www.atomasevic.com/cv.pdf
    Email: atomashevic@gmail.com

  -----

Hi, I'm currently working in academia as an Assistant Professor of Quantitative Methods at the Social Sciences Department and I'm working on transitioning to the industry. I obtained my PhD in 2019 and have been working on computational and quantitative methods since then. My latest project is the R package transforEmotion, which is used for sentiment analysis and emotion detection in text, images, and videos: https://github.com/atomashevic/transforEmotion.

I have experience with data science/ML projects in academia, as well as consulting for NGOs and companies. I'm open to remote work possibilities as a full-time data scientist or research scientist. I'm eager to work on challenging technical or business problems. Please reach out if you think I might be a good fit for your team!

I'm the listener who forwarded these screenshots to Cal.

They come from Gödel's Philosophical Notebooks, specifically volume two which is about time management (https://www.degruyter.com/document/doi/10.1515/9783110686586...).

Notebooks provide fascinating insight into how Gödel dealt with issues of focus, productivity, producing meaningful work, dealing with distractions and similar topics Cal and many others often talk about.

These notebooks are from 1934 to 1955, the period in which he struggled to produce impactful work after incompleteness theorems.

I get that you're the only user of the tools you're creating, but can we at least get some more info on `soc` and `burg`.

For starters, details about that wonderful colorscheme you're using :)

Thanks, I quickly found what I was looking for!

Discord sounds like something I would enjoy, I was heavy IRC user back in the day. But I never found an interesting server.

Any recommendations for servers focusing on topics like AI-ML/Python/Math,probability, statistics?

I notice similar in my case and case of my friends (born between 1985-1990).

I've seen fair share of new services and social networks being branded as "Facebook killer" in last 10 years but for me group texting was the thing that replaced the main use case of Facebook and Instagram - sharing important/funny/interesting moments and stuff with friends and family.

Back in the day I could use IRC or some messengers for this, but Facebook was the first to enable this at scale. Afterwards messaging apps (Viber in case of my country but I guess Telegram/WhatsApp for the rest of the world) made this more convenient but also more intimate/rewarding.

We don't have mutes or bans in our group texts. Discussions, even political ones, don't blow up like on Twitter. There's less incentive for things to go crazy when you're exchaning intersting/important stuff with people you know and care about.

Also, before Facebook I created and administreted phpBB forum for the same group of friends. We were in late teens, early 20s at that point so the main topics were "What are you listening too?", "What are you drinking tonight?", "How was Saturday night for you" (but also topics around literature, comic books, movies, video games). I get similar vibes from our group texts that I had on that forum more than 15 years ago.

I agree and I guess I wasn't clear in my comment. It didn't work for me, but for other people and/or use cases it might, no doubt.

Seeing others quit TfTs and intricate zk systems on HN helped me realize that I should probably do that too, again given my situation. So, just sharing my experience, not trying to preach :)

Go out of the way to imerse yourself with the subject. Switch locations, find a cool place to study (park, coffee shop, library, ...) and take notebook/books with you, put a particular type of music on the headphones. Make studying a special, different time of day/week which is contrast with other activities

As others said, you need to be fascinated and willing to dive deep, obsess (in a positive way) over details, tinker, play around with code/materials/math...

I've spent a good (unproductive) part of 2022 tinkering with notes/TfTs and I'm stuck with "optimizing" note-taking for more than few years (started with zettelkasten and Roam Research I guess).

I'm stopping with all of that - zero results and tons of wasted time (and fair ammount of subscription money). If I spent all that time tinkering with LaTeX at least I would leveled up my skills.

Judging from comments on HN, I'm not alone and quite a few of us went down weird notetaking rabbit holes. I'll just stick with few scratch plaintext files and handwritten legal pad notes to capture stuff as I work and then transfer/delete stuff at the end of the day without any special system behind the whole process.

I just switched to Bitwarden after seeing it recommended on HN a bunch of times. Bought a subscription right away.

I previously stored everything in Firefox, transfered it easily to Bitwarden. Linux app seems to work fine, tested in Firefox, Chrome, Android phone, smooth transition.

The only thing that I've noticed is that you have to change existing passwords manually by editing records in the vault, the Firefox extension does not prompt you to update password once it detects a succesful login with another one.

Great, give "Doomsday Clock" limited series a chance after you finish with Watchmen, it's an interesting crossover with mainstream DC Universe.

Congratulations and thanks for sharing! I'm on the verge of making the same step. Hopefully, I will have the same kind of respect from friends and coworkers.

The extension on VS Marketplace:

http://marketplace.visualstudio.com/items?itemName=atomashev...

I completely agree with the conclusions, GitHub repositories can really be the focal point of scientific publishing in years to come.

I always wondered why it took so long to implement math. There are better implementations out there for sure, it's weird they waited for so long and implemented something that is below the standard of alternatives.

The next step is to render citations from .bib files. I hope they get that right in the future.

Trying to solve a large problem set from a statistics textbook. If I can finish them on time, I'll probably tinker with the simulations around those problems in R and try to build more intuition about hard concepts.

The rest is family time. I can perhaps sneak one or two papers to read that I really look forward to.

Thanks for sharing! I made mistake in my comment when I called closed Stackexchange sites "failed" because the core users might have migrated to some other platform or formed another form of a meaningful community.

We didn't focus on individual contributions, only on community-level metrics. The reputation metric we used rewards frequency of contributions (questions, answer, comments) and that turned out to make a difference in building a strong core.

Also, core is not static, sustainable communities have the ability to "pull" new users into the core, so new, enthusiastic users are not discarded by the existing core.

If you are interested in individual efforts and user archetypes take a look at another work: https://arxiv.org/abs/1709.05111

I'm sorry to hear that.

One of the reasons we focused on smaller niche SE communities is because Stack Overlflow became completely "another beast" where the processes of valuing contribution and building reputation work in a very different way (or don't work at all) on a large scale.

Hi HN,

Here's a paper we've been working on for some time. Essentially, we looked at differences in failed vs successful SE communities on the same topic (e.g. failed vs launched Astronomy SE) and tried to identify what made them different in the early days (first 6 months).

I hope it may be useful/interesting to those interested in (q&a) community building.

Here's the abstract:

Knowledge-sharing communities are a fundamental element of any knowledge-based society. Understanding how they emerge, function, and disappear is thus of crucial importance. Many social and economic factors influence sustainable knowledge-sharing communities. Here we explore the role of the structure of social interactions and social trust in the emergence of these communities. Using tools from complex network theory, we analyze the early evolution of social structure in four pairs of StackExchange communities, each corresponding to one active and one closed community on the same topic. We adapt the dynamical reputation model to quantify the evolution of social trust in these communities. Our analysis shows that active communities have higher local cohesiveness and develop stable and more strongly connected cores. The average reputation is higher in sustainable communities. In these communities, the trust between core members develops early and remains high over time. Our results imply that efforts to create a stable and trustworthy core may be crucial for building a sustainable knowledge-sharing community.