Come to think of it, LLMs are a raging hit partly because they are never pitched to be perfect or personified (unlike other AI tools in the past). Any newly found shortcoming of LLMs is more a caution to the user than an issue to its makers. Brilliant!

To all the prospective PhDs in/from India, before you sign up for a PhD position in the US, please carefully consider if you would do better (PhDing) in India. I sketched my experience in this blog vihari.github.io/why_india/. Feel free to send me your comments 😃

🚨 India prohibits coaching centers from enrolling students below 16 years of age.

IKDD congratulates Vihari Piratla (IIT Bombay) for his work on tackling the critical challenge of handling distributional drifts in ML deployment, and making impactful contributions in generalization, adaptation, and evaluation of models under changing domains.

I wrote a short summary of AI Alignment that you may find handy given the many (somewhat needlessly introduced) terms. I also included some personal thoughts. Happy to hear your thoughts. gist.github.com/vihari/0d2af83…

This looks amazing: capturing carbon from seawater and generating hydrogen as a byproduct. Given the rate the climate is changing I am sometimes wondering why we are not pouring billions into these projects and put ITER / LHC on the backburner for a while. newatlas.com/energy/equatic…

I reckon AGI debate meaningless. Intelligence is demonstrated through thinking. And thought is a function of knowledge. AGI in a nutshell is to wholly amass world’s knowledge. Knowledge, let alone the digital form, no matter how vast is limited. And so is Intelligence (AGI).

From 30 off 30 required for SA to winning by 7 runs. What a match!! Loved the way India kept their nerves.

The blog reinforced my belief that happy/content people make less "successful" researchers.

Nepali mountaineer Chhring Namgel Sherpa on the summit of Mt K2 (8611 m) in #Pakistan.

Sherpas are just built different. Exercising on the K2 summit without any oxygen support, insane!!!

In her #ACL2024 keynote, Sunita Sarawagi seems to diss the Ersatz Natural Science of LLM Emergent behaviors.. 😱 (see also cacm.acm.org/blogcacm/ai-as… )

Doing good science is 90% finding a science buddy to constantly talk to about the project.

I met PM Modi several times w/the same discussion: What India needs is make failure honorable; smart people tend to go to prestigious employment in a bureaucratic system instead of building startups. He said that in Gujarat people who failed wore a literal badge of honor.

Congrats Abhijeet Awasthi for the well deserved award. IITB is a wonderful place to do PhD, and Sunita is an amazing advisor 😃

Amazing from Google DeepMind! Congrats to the whole team.

Amartya Sen is such an inspiration!

Sets an interesting baseline in learning robust models when causal attributes are known and can be controlled. The gains are consistent and significant.