Saptarshi Ghosh

Postdoc at Institute of Neuroinformatics(INI)....Conplex Researcher gone Silicon Neuro ...
#LifePhilosophy #BeLikeaGel

[Saptarshi Ghosh @sapta15]

Check out my #WebSpace for recent articles and news @
https://t.co/zUHMRmJGI3

[sphinx @protosphinx]

Sir, they have blocked access to the very AI models you said we should share instead of build

[Rohan Paul @rohanpaul_ai]

This week, the Linux kernel project finally created a formal, project-wide policy explicitly allowing AI-assisted code contributions, as long as developers obey strict new disclosure requirements.

Torvalds’ view, which gives this policy its main philosophical shape, is pretty direct: AI is just another tool. Developers submitting garbage code are not going to be fixed by more documentation, so the kernel should hold people accountable instead of trying to control the software they use on their local machines.

It is a practical and reasonable line to take, especially compared with the panic in other parts of the open-source scene.

You are the one on the hook now. If Claude introduces for example, a race condition in the block layer and you approve it, the patch carries your tag, not the model’s.

The Signed-off-by line is the certification for the Developer Certificate of Origin, and the latest policy makes it explicit that only humans can legally add it. AI agents "MUST NOT"

The open-source community is currently getting overwhelmed by what people are calling "AI slop."

e.g. the creator of cURL closed bug bounties after a flood of hallucinated code, tldraw began automatically closing external PRs to defend itself, and projects such as Node.js and OCaml have seen huge, >10,000-line AI-generated patches

[Saptarshi Ghosh @sapta15]

@PhysInHistory I wonder if he would have gotten away with his radical ideas in today's academic atmosphere and norms!

[Saptarshi Ghosh @sapta15]

😂

[Retraction Watch @RetractionWatch]

‘The fraud was not subtle’: Chemist blames students after ten papers retracted.
https://t.co/FlP95sqlfx

[Saptarshi Ghosh @sapta15]

@ChoptecOfficial @geerlingguy @Raspberry_Pi @threadreaderapp unroll

[Saptarshi Ghosh @sapta15]

@soledad_gcogno @forskningsradet @KISNeuro Congratulations. Well Deserved!

[Manu Awasthi @mnwsth]

Flooding of roads after a 10 min rainfall, among many other such examples, is a consequence of lot of investment in infrastructure and not enough in education.

[Saptarshi Ghosh @sapta15]

@DelhiAirport Don't have to look abroad! Just look at @BLRAirport how its done!!

[Saptarshi Ghosh @sapta15]

International departure at T3 @DelhiAirport is a mess! Tiny trays, outdated XRays, and zero crowd control—truly frustrating. Hats off to the staff and security doing their best with two and half hands tied to their backs. #TravelWoes #DelhiAirport #TravelAlert #T3

[Saptarshi Ghosh @sapta15]

@DelhiAirport Don't have to look abroad... Look at @BangaloreAirport or others

[Saptarshi Ghosh @sapta15]

@DelhiAirport And What the hell those metal detectors do? It's not the 2000s anymore. Every single person gets a red light and a person checks them with a hand wand. It's probably fine to be domestic but in international 🤨🙄🙄🤌🤌

[Saptarshi Ghosh @sapta15]

@DelhiAirport And apparently @DelhiAirport gets selected as best Airport 6 consecutive times in Asia region 🙄🤌

[Saptarshi Ghosh @sapta15]

@IndiaPostOffice Frustrated with a parcel I was promised would arrive in 10-12 days. It’s stuck in transit with barely any tracking updates. This delay is unacceptable for a domestic, low-weight parcel. This is very disappointing...

[Saptarshi Ghosh @sapta15]

@HDFCBank_Cares @HomeLoansByHDFC @HDFC_Bank At the same time, I really appreciate their (atleast from some of them) eagerness to help despite having very limited information.

[Saptarshi Ghosh @sapta15]

Its been a long struggle but appreciation is due where deserved. Thanks @HDFCBank_Cares @HomeLoansByHDFC @HDFC_Bank for processing my transaction. Specially heartfelt gratitude to staffs of Jalpaiguri and Siliguri Loan Branch who helped us all the way to the end. #ThankyouHDFC

[Saptarshi Ghosh @sapta15]

@HDFCBank_Cares @HomeLoansByHDFC @HDFC_Bank Same feedback as before, (I) train your staff well, specially in non metro small city branches. (II) Try to process all NRI related transactions from a central location (III) reduce staff rotation (I went through 4 BMs in last 1.5 years)

[Saptarshi Ghosh @sapta15]

@mohakmangal saw your take on NEET today! alongside UG and PG, Higher Education and Research is also an important pillar for our technological progress! Please cover ANRF if possible. It is important for researchers who are hoping to return.

[Arindam Ghosh @ArindamPhysics]

ANRF is an opportunity to function differently. It will be a pity to lose it. The research funding modality in India needs serious overhaul.
https://t.co/OOwfB5jy3P

[Manu Awasthi @mnwsth]

Just thought of doing a random comparison, of a top rated Chinese University (Tsinghua) and top rated Indian University (IISc).

I am not picking on any institution, but trying to make a comparison. Generally, the top performing institutions are the best funded ones. At least according to the levels of funding in the country under consideration.

I am going by the data available on Wikipedia which might not be completely correct, but allows us to get ballpark numbers.

Compared to IISc, Tsinghua has:
- 6.8x the faculty
- 5.75x PhD students.

Tsinghua also has a USD 6B endowment. I don't think IISc has one. And even if it does, it will be very, very small (to the extent that the government rules allow)

In addition, Tsinghua graduates 10s of thousands of UGs and PGs every year, which IISc doesn't, but lets leave that aside for a moment.

This didn't happen overnight. It took a long time (decades), a lot of effort, and a lot of money to make this happen. The country decided that it is in their long term interests to fund research and universities within the country. They got the best people to work in these places, gave them the facilities they needed -- funding, spaces, equipment, incentives to get research going. And they kept doing it for a long, long time without expecting to see returns right away. I personally know of cases where top universities in China made such good offers to excellent candidates that no sane person could reject them.

And the results are there for everyone to see.

BTW, Tsinghua is just one example. There are many such universities in China that have similar levels of funding and research output.

Our top institutions do the best they can. But, end of day, they are shackled by the lack of vision and commitment of successive governments to develop the country as a research powerhouse. And then there are myriad bureaucratic rules on top of that make it even more difficult.

[Valerio Capraro @ValerioCapraro]

Just finished reviewing several hundred applications for a very selective PhD program (3% acceptance rate). I’d like to share a few tips on how to write a strong research statement and avoid common mistakes.

First, the sad reality: we receive many statements that seem to be written by ChatGPT. Using ChatGPT to refine your writing is fine, but overrelying on it is a recipe for failure, because ChatGPT produces just average outputs. If you’re applying for a selective program, you don’t want to be average, you want to stand out. So:

Tip 1: Be unique, show what unique contribution you can bring.

Many candidates focus solely on how a PhD would advance their career. This is a mistake. You don’t have to talk only about your interests but also about ours. The trick is to align your interests with those of the institution. You have to argue why you would be an added value for the institution.

Tip 2: Show how you would benefit from the program AND how the institution would benefit from you.

Many applicants write impersonal statements that could be (and probably are) submitted to many institutions. This is another common mistake. If you want to work with us, show us why. Name specific faculty members you wish to collaborate with and reach out to them before applying. But don’t send impersonal emails to dozens of professors: we appreciate genuine interest from prospective students, but we don’t like to be spammed.

Tip 3: Explicitly name potential supervisors and contact them beforehand. But don’t spam

Many candidates write super long and detailed research projects, while others write just a few vague lines about their research interests. Neither is optimal. Your project is likely to change a lot during your PhD, so there’s no point in writing a detailed research project. On the other hand, we’d like to see that you have an idea of the open problems in the field and that you have a strategy to address them.

Tip 4: Focus on one important research question and give us an idea of how you plan to address it.

I hope this is useful. Feel free to ask questions.

Good luck with your application!