
PhD, @MIT. BTech, @iitmadras. AI, ML, Climate, Upskilling, Education. Views are my own.
Bengaluru, India
Joined September 2009
- Tweets3.4K
- Following789
- Followers3.1K
- Likes5.1K
Before going to IIT, I thought I was 1. But then I saw real 1 at IIT, then to cope, I went the route of 2. That took me ahead and now I am a full fan of 2.
I think only thing that will make sense is allowing people to submit only if they have previously reviewed as per community standards. Being reviewer in a venue where you are an author with limited slots is CoI
It really is pointless and unnecessary to allow external reviewers to use AI as they want. There are obvious better alternatives.
By now, everyone should already be using AI to check their work before submitting. So what's the point of reviewers meat proxying to send more AI-generated feedback? Seemingly none.
What happens if we ban AI use in peer review? The main risk is to the imagined role of peer review as a basic soundness filter. But this is much better addressed by the venue having a centralized approach that uses validated AI tools to check that a paper is what it says it is than by individual reviewers feeding papers to LLMs using basic prompts. Such checks might include things like whether the results can be reproduced from the data/code, whether claims can be traced back to some provided evidence, whether there are TODOs left by an LLM to replace table results with non-fake data, etc.
Some of these can be used to determine if a paper should be desk rejected. Anything that is more controversial can be surfaced for reviewers.
This will shift what reviewers pay attention to, but this can be addressed by: 1) making clear to reviewers that their role is to identify what these checks miss and to weigh them, 2) being intelligent about how you select the set of LLM checks that are surfaced. See e.g., our work on learning LLM-implemented checklist that avoid adding more bias than they reduce variance. arxiv.org/pdf/2610.28666
What happened with ICLR being permissive with reviewers using AI freely is seeming pretty inexcusable.
Now I understand why the PM has so much support in Gurjar Pradesh. Karnavati is a top class world city. Wide roads, modern amenities, so many open spaces and things to do. Just if the weather was good…. Had a fun time at Kankaria Lake, Science City, Riverfront etc
Deepak Subramani retweeted
Annachi takes on AI coding 🥲 No AI coding took on Annachi 😂 Catch Retired NOT Hurt in Chennai on Nov 21 and Bengaluru on Dec 5 - the last two shows of 2026.
Vaanga Makkale! ❤️ 🇺🇸 🇨🇦 🇦🇺 Coming soon 🙌🤗
For almost all X and LI interactions, I feel like saying Suraj’s dialogue about Murthy Annan: “He didn’t use the elephant, nor did he strike the king. Yet, how confidently he made the move and say impresses me” from Jailer 2.
Actually there is another angle to this story. If rich states keep all their winnings and grow, then it is a problem for them unless border control is put in place. Redistribution and quick growth of poor states is essential.
You know that VB is fast because they get priority over other trains. Of all trains are VB, nobody gets priority on our congested network @cvkrishnan sir
Deepak Subramani retweeted
Yup, it's the new form of performative busywork.
“At the end of the day, most people just want smart, hardworking people that are fun to work with and aligned with the mission. Of course there needs to be alignment in skillset, but I think AI is the great democratizer that makes that less relevant” AGREE
I used to refuse to do Leetcode interviews out of principle. I thought they were stupid, and I'd rather build. In hindsight I left a lot of money on the table by being so principled, but I never struggled to find work. I even took a year and a half off once to travel the world, and when I decided to get back to work it only took me 3 weeks to secure a remote job that allowed me to stay abroad in Asia (this was before COVID when remote was still uncommon)
Then in '24 it was clear to me that the market was no longer what it used to be, and was only going to get worse (don't forget at the time everyone was claiming this is all temporary). I swallowed my pride, moved to SF, and grinded my Leetcode like a good little monkey. 500 questions later I secured an offer at Meta. I felt like I'd slid into the garage just before the door shut.
Now in '26 coding has been automated, but most companies still do coding interviews. Never before have coding interviews been more removed from reality, and I've personally lost all motivation to prep for them. Many if not most companies will probably inevitably phase them out within the next year or two, but then the question becomes - what will they replace them with? If companies are replacing them with AI-enabled HackerRank/CodeSignal interviews, I wouldn't consider that an improvement based on those interviews I've seen.
IMO all interviewing is flawed. At the end of the day, most people just want smart, hardworking people that are fun to work with and aligned with the mission. Of course there needs to be alignment in skillset, but I think AI is the great democratizer that makes that less relevant than ever. If I were running my own company, I'd just find those people and give them work trials. All this monkey hoop jumping does nobody any favors, and is frankly demeaning. It's always been that way and we've all known it, but now it's finally gotten so absurd that the tide is finally shifting.
Maybe because I’m an ESL speaker, I find agent English easier to understand and not the rewrite as given by @sh_reya
I guess this is the only way going forward. Shortlist on the basis of resume is impossible. Direct walk in is as good as a shortlist from AI generated resume
Deepak Subramani retweeted
I think voter apathy or long weekend is a very small problem compared to the data errors that have a very high probability and plausibility. The attack on hard-working tech bros who may have taken off to a resort after voting at 7 am is uncalled for.
Reminder to everyone about my post after lok sabha turn out. Bengaluru SIR was very much needed.
I agree
Winter is coming for publish-or-perish research.
People often ask, what’s the point of doing traditional research anymore.
I think we are entering a new era of research.
The era of doing research for publications and incremental improvements is fading. The era of research that creates real-world impact will take over.
We went from Agentic AI yeay to extreme fear in 2 days?
Helpful prompt about what to do if the internet goes down due to a global cyberattack/AI swarm.
This seems less insane with every passing day.
I also bought myself the SAS Survival Guide, which is essentially a bush manual for almost every imaginable outcome.
Money
- Keep cash at home, about two to four weeks of essentials, in small bills.
- Write down your bank, credit card and insurance account details and phone numbers on paper.
Information and contacts
- Print a contact list with family, close friends, neighbors, doctors and the kids’ school.
- Agree on a family meeting point and plan in case phones are down, such as “if we can’t reach each other, meet at X by Y time.”
- Export and print or save offline anything business-critical.
- Download offline maps
- Keep paper copies of passports, IDs, insurance cards and prescriptions.
Communication
- Get a battery or hand-crank AM/FM/NOAA radio, which will be your main news source.
- Consider a pair of walkie-talkies or GMRS radios for the family.
- Keep a landline-style backup in mind. Most “landlines” now run over the internet, so don’t count on yours.
Power
- Keep power banks charged, plus a solar charger or a larger portable power station.
- Have flashlights, headlamps and spare batteries.
Water and food (aim for 2 weeks)
- Store about 1 gallon of water per person per day. In a high-rise this matters most, because building water pumps need electricity.
- Stock shelf-stable food the kids will actually eat, plus a manual can opener.
- Keep a camping stove with fuel for outdoor or ventilated use only, or no-cook foods.
Health
- Keep at least a 30-day supply of any prescriptions.
- Have a first aid kit, OTC basics like pain relievers in children’s doses, and hygiene supplies.
Home and offline life
- Learn how to manually open anything electronic, such as smart locks and the building’s garage or entry.
- Keep a paper copy of the building’s emergency procedures and super’s number.
- Download offline entertainment and learning material for the kids: books, games, art supplies.
- Get to know a few neighbors, since local community becomes your network fast.
Digital hygiene (helps before and after)
- Back up photos and documents to an external drive kept offline.
- Use strong, unique passwords with 2FA, and keep printed recovery codes for key accounts.


















