Pinned Tweet
Happy to share: My 20 hour course on Practical Machine Learning with @TensorFlow 2.0 is now live on @nptelindia & @YouTube . It's a joint course between @Google & @iitmadras. Thanx @divy93t @pankaj @ravi_iitm @random_forests & @lmoroney for support. Strong TF interest in India.
Ashish Tendulkar retweeted
Some news: I'm starting a company in the AI for Science space.
We are training foundation models and improving them for multi-agent scientific discovery.
If you're looking to join an early-stage startup as a Research Scientist or Research Engineer, I'd love to chat. My DMs are open!
Ashish Tendulkar retweeted
We're excited to announce the 2026 recipients of the Google PhD Fellowship!
148 new PhD students across 38 countries (including 9 from India), with $13M in support from @Googleorg: goo.gle/phd-fellows-2026
Huge congratulations @OfficialLoganK
1
174
Ashish Tendulkar retweeted
Welcome! We are very excited
John -> Core Automation coreauto.com/
After Microsoft (hunch.net/?p=13764006), I was attracted to a place which was very willing to try new things and exceptionally enabling in doing so.
Ashish Tendulkar retweeted
Announcing a major milestone achieved by my team @GoogleDeepMind: we have proven key security properties of the machine code underlying the Signal messaging protocol. This is the first time the actual production code of a widely deployed protocol has been formally verified – by any method, with or without AI – and at this depth.
More than a research breakthrough, it’s a prime example of how frontier AI can be a force multiplier to counter the speed and scale of modern cyber threats.
Formal verification offers a higher standard of security; however, it’s extremely complex and time consuming, which is why in many cases it’s considered to be impractical. Similar efforts could take teams decades of combined effort. We did it in a month’s time.
Using AlphaProof Code, Google DeepMind's new agentic framework for formal verification that runs on LLMs including Gemini Carbon, we verified the Signal protocol down to the actual code running on devices. This milestone lays the groundwork for a future in which vital systems, from medical device firmware to power grid software, ship with mathematical proof that their protections work.
Reaching that future will take continued research from our community. In that spirit, we're open-sourcing our formal specifications and verification artifacts so others can build on this work.
On a personal note, this moment is of deep significance to me. I started my own research journey more than two decades ago as an intern working on checking the consistency of complex software systems. Seeing this grand challenge made practical with modern AI is extraordinary.
I'm extremely proud of my team members: Swarat Chaudhuri, Andrew Ferraiuolo, Anton Kovsharov, Tej Chajed, Sam Schlesinger, Karn Seth, Jonathan Katz, and Charles Sutton for taking on this grand challenge and delivering, and grateful to the Signal team for their partnership.
deepmind.google/blog/formal-…
Ashish Tendulkar retweeted
My article on #Nvidia honeytraping of #JensenHuang
From the information that is surfacing in this issue, I won't be surprised if #JensenHuang is just a facade and real dealmaker is #POTUS
Ashish Tendulkar retweeted
We also investigate mathematically why reading trained weights is difficult, examine the agents' hypotheses against the actual network, and discuss what this tells us about Weight-Space Programming.
Ashish Tendulkar retweeted
Excited about the launch of Gemini for healthcare! Bringing the best of our models into clinical workflows
Today, we’re announcing Gemini for healthcare.
This newest addition to our suite of purpose-built industry solutions extends the Gemini agent to health systems and health plans.
Learn more → goo.gle/4zf0s11
AI detectors are required everywhere now.
In about 3 weeks of release, SnapJudge -- our AI detector for scientific text -- has checked over 25,000 submissions so far!
Kudos to the Yixuan (@huang13258) for building this... encouraging to see so much demand for free open-source AI detectors.
4
298
Ashish Tendulkar retweeted
I am looking for a 6-month research intern (paid) to work with me on exciting problems in Agentic AI at @LatentForceHQ.
Full-time, on-site (Bengaluru). People with demonstrated prior research experience will be prioritised.
Apply if interested (Link in the first reply).
I'm delighted to partner with @LaudeInstitute (along with my @Sanjay_Ghemawat, John Hennessy, Dave Patterson, Ben Gomes and @andykonwinski) to support funding of open science moonshots!
anyone who is close to research knows how broken the incentives are. doing open science at the frontier is harder than ever, compute costs alone price most teams out. we kept wondering what might happen if we give the right resources to the right researchers at the right time?
laude.org/updates/moonshots-…
Simran would be a great addition to any team be it frontier labs or an academic setup.
I'm on the industry job market this year!
I'm a PhD student at CMU LTI. My work spans post-training, evals, alignment and multimodality!
Please reach out (DM) if you are hiring and think I'd be a good fit, I'd love to chat :)
Website: simran-khanuja.github.io/
CV: simran-khanuja.github.io/ass…
Irrespective of AGI or ASI, AI remains a useful tool and has horizontal appeal like productivity tools. Beyond that it's fashionable to label everything as AI and that bubble will surely burst.
Ashish Tendulkar retweeted
Shame on @DarioAmodei / Anthropic. I never distilled Claude—rollouts came from a strong open-source model; Claude just helped submit jobs and run analysis. Still got tagged for "training a competitor," and now anything touching RFT, SFT, or training larger models gets me blocked, Opus 5.5 downgraded to Opus 5, or the session killed outright. You built your models on the world's data, then treat paying users doing research automation as the enemy. Pathetic.
Ashish Tendulkar retweeted
☎️ Calling all students:
You can now access hundreds of textbooks directly in Notebook, for no charge. From publishers like S Chand Publishing, OpenStax, and Open Textbook Library— browse, select and upload them as sources to ace all of your upcoming exams (no more excuses, ok?)
Amazing work @acharyaamrit . Congratulations.
I was 29 when we started Zetwerk. My wife Sharanya and I had just returned from the US to build our lives in India. Last month, we welcomed our daughter. Last week, we took the company public. It is yet to sink in.
When I moved back, my father, a professor, was sceptical. I had walked away from the traditional Indian dream of a job abroad. But our story is also that of today's India, where my cofounders and I, 20-somethings from small towns, without family connections, can raise capital, build a business, take it global and take it public, all within a decade.
The world today is on the cusp of enormous progress: defence modernisation, energy transition, space exploration and AI buildout. Much of this progress faces a common bottleneck: building physical things in the real world.
But the manufacturing systems of the past won't build the future. Manufacturing must modernise to catch up with the industries it serves.
At Zetwerk, we always believed there is enough depth in Indian manufacturing whether it is within the small workshop in Peenya which is great at tooling or it is in the new plants we have set up to drive scaled production. Our approach combines real-world engineering with a digital stack to unify ~7000 suppliers, 25+ owned facilities, 1000+ customers and 3000+ employees into a single layer that operates with the predictability of a world-class factory, standardising quality and delivery. We call it the Universal Factory.
Our ambition is to build a world-class industrial institution bringing together two distinct worlds: software and hardware.
Think of your day today. You may have made coffee, worn headphones, taken the metro or a flight, or opened your laptop. There is a little bit of Zetwerk in all of these.
We want our customers placing complex, multimillion-dollar orders to have an experience that is "boringly predictable" - simple, transparent and reliable. We have over ₹15,900 crore in revenue, more than 80% repeat customers and a manufacturing order book of ₹13,000 crore. We've come a long way, but also have a long way to go.
On IPO day, my father stood beside me as we rang the bell at the National Stock Exchange. He is not sceptical any more :)
Someone wise once said “You make your own reality. And once you've done it, everyone's of the opinion it was all so obvious”. That is how the future works. Before something transformative exists, people call it improbable. Afterwards, they call it inevitable. This is the spirit in which we are building Zetwerk: with a relentless commitment to the future.
Great industrial institutions that came before us, like GE, L&T and Tata, helped make the modern world possible. Today, Indian manufacturing is having its moment, and we intend to build what comes next. A big thank you to our investors, customers, employees, their families, our bankers and partners who helped us get here.
We are committed to building a world-class industrial institution out of India.
1
161
Ashish Tendulkar retweeted
Frontier labs are advancing the capabilities in AI for biology but preventing the good guys (scientists, academia, non-profits, health agencies,...) to use, study and understand it with widespread but shallow guardrails that the bad guys are bypassing easily (blocked me to proofread this tweet for example 😅).
This will automatically create massive assymetry and risks in the coming years, just like in cybersecurity, especially now that frontier labs are building wet labs (HF cybersecurity scenario for biology when?).
This is why we created Carbon-X, the best open-source DNA model by midtraining Qwen3-14B on 300B tokens of DNA, English and interleaved data. We also used a hybrid tokenizer that tokenizes English with BPE and DNA with a 6-mer tokenizer.
As a result, Carbon-X outperforms all open language models we evaluated on five benchmarks of DNA understanding and generation. This is a 27-point improvement over Qwen3-14B, showing how much a language model can gain from DNA training. It is also API frontier level at 1,800x less cost than GPT 6.1 Sol!
We hope Carbon-X helps researchers explore DNA in new ways. Releasing it openly means anyone can study its capabilities, limitations, and risks. Let's make the AI biology frontier open-source to make it safe and impactful for the whole world (not just a few companies).
Apache 2.0 Model on @huggingface: huggingface.co/HuggingFaceBi…
Ashish Tendulkar retweeted
A unique concept by a group of renowned musicians. A pitch event for tech solutions in Indian Art. If you are a art-tech enthusiast, please apply here. I'm very curious to know what gets built. Apply here - tinyurl.com/yc5n96nk
I'm not a "Shark though", may be a dolphin :)


































