Anoop Kulkarni
@anooprkulkarni.bsky.social
PhD in computer vision. Interests in Causal inference and AI in healthcare.
Just barely through the first hour - but I liked the intro. Maybe AI will solve "artificial medicine" :) Both real intelligence and real medicine look like a distant dream!
October 31, 2025 at 11:01 AM
Just barely through the first hour - but I liked the intro. Maybe AI will solve "artificial medicine" :) Both real intelligence and real medicine look like a distant dream!
I think the most natural place would be to have some additional material on CI included in a stats course in curricula but also everyone in day to day life talks of "root cause of this is that" etc but have no formal background/ exposure to wade through this!
October 21, 2025 at 6:43 PM
I think the most natural place would be to have some additional material on CI included in a stats course in curricula but also everyone in day to day life talks of "root cause of this is that" etc but have no formal background/ exposure to wade through this!
Nice suggestions there and the backdrop - wow!
I am a practitioner and a consultant in healthcare and in industry every time I have to start conversation from what causality is for every project and build upwards.. Frankly it is exhausting starting all over from the Pearl's book of Why each time!
I am a practitioner and a consultant in healthcare and in industry every time I have to start conversation from what causality is for every project and build upwards.. Frankly it is exhausting starting all over from the Pearl's book of Why each time!
October 21, 2025 at 6:11 PM
Nice suggestions there and the backdrop - wow!
I am a practitioner and a consultant in healthcare and in industry every time I have to start conversation from what causality is for every project and build upwards.. Frankly it is exhausting starting all over from the Pearl's book of Why each time!
I am a practitioner and a consultant in healthcare and in industry every time I have to start conversation from what causality is for every project and build upwards.. Frankly it is exhausting starting all over from the Pearl's book of Why each time!
Thank you. Will take a look at the recording. Wanted to join but was the Diwali time here in India that exact hour and was in a different festive zone!
October 21, 2025 at 4:22 PM
Thank you. Will take a look at the recording. Wanted to join but was the Diwali time here in India that exact hour and was in a different festive zone!
AGI is a marketing gimmick to say as if AI is done deal.
AI remains unsolved and will for some time to come, recent advances notwithstanding. Make no mistake I am in awe of what has been achieved and yet the goals of AI sinces 60s havent really changed. Have they? That's just my take.
AI remains unsolved and will for some time to come, recent advances notwithstanding. Make no mistake I am in awe of what has been achieved and yet the goals of AI sinces 60s havent really changed. Have they? That's just my take.
October 17, 2025 at 5:04 AM
AGI is a marketing gimmick to say as if AI is done deal.
AI remains unsolved and will for some time to come, recent advances notwithstanding. Make no mistake I am in awe of what has been achieved and yet the goals of AI sinces 60s havent really changed. Have they? That's just my take.
AI remains unsolved and will for some time to come, recent advances notwithstanding. Make no mistake I am in awe of what has been achieved and yet the goals of AI sinces 60s havent really changed. Have they? That's just my take.
The "blue machine" book is amazing and learnt so much from it!
October 16, 2025 at 3:55 PM
The "blue machine" book is amazing and learnt so much from it!
I am hopeful and I do my bit to keep people informed of various celestial events.. sure can't see the milky way but I do share the joy of seeing Venus, jupiter, saturn, mars and the moon and the eclipses and indeed the many constellations.. stargazing is better than phonegazing :)
October 15, 2025 at 9:29 AM
I am hopeful and I do my bit to keep people informed of various celestial events.. sure can't see the milky way but I do share the joy of seeing Venus, jupiter, saturn, mars and the moon and the eclipses and indeed the many constellations.. stargazing is better than phonegazing :)
without the chatGPT, this was a great abstract of the paper. Thanks - will read more about causal consistency!
October 14, 2025 at 6:50 PM
without the chatGPT, this was a great abstract of the paper. Thanks - will read more about causal consistency!
Wow - many congratulations! I keep traveling to London - so guess I will get to meet you after those initial days of interaction at NMA
October 13, 2025 at 5:24 PM
Wow - many congratulations! I keep traveling to London - so guess I will get to meet you after those initial days of interaction at NMA
Great! Thanks !!
September 26, 2025 at 3:20 PM
Great! Thanks !!
Behind a pay wall is it?
September 26, 2025 at 12:46 PM
Behind a pay wall is it?
All you need is data :) and not the data generating process!
September 26, 2025 at 6:20 AM
All you need is data :) and not the data generating process!
I struggle to communicate to the largely AI/DS/ML community here that CI is central.. not data.. its an up-hill but I like the challenges!
September 26, 2025 at 6:18 AM
I struggle to communicate to the largely AI/DS/ML community here that CI is central.. not data.. its an up-hill but I like the challenges!
No escape route!
September 12, 2025 at 5:29 PM
No escape route!
These days (in Python) my goto libraries are pyWhy (doWhy), causal-learn and your library 'causality'.
Everything else, as need be, am building my own little one, but its not generic one in a long way.. more tailored to my needs..
Everything else, as need be, am building my own little one, but its not generic one in a long way.. more tailored to my needs..
September 12, 2025 at 3:38 PM
These days (in Python) my goto libraries are pyWhy (doWhy), causal-learn and your library 'causality'.
Everything else, as need be, am building my own little one, but its not generic one in a long way.. more tailored to my needs..
Everything else, as need be, am building my own little one, but its not generic one in a long way.. more tailored to my needs..
I am sure you have come across the foundational work in this area by Janzing et al (2012 ish) but sure its a bit dated.
I too am looking for a comprehensive Python library that implements information-geometric causal inference. There are a few that have it in bits and pieces - but none fully.
I too am looking for a comprehensive Python library that implements information-geometric causal inference. There are a few that have it in bits and pieces - but none fully.
September 12, 2025 at 12:42 PM
I am sure you have come across the foundational work in this area by Janzing et al (2012 ish) but sure its a bit dated.
I too am looking for a comprehensive Python library that implements information-geometric causal inference. There are a few that have it in bits and pieces - but none fully.
I too am looking for a comprehensive Python library that implements information-geometric causal inference. There are a few that have it in bits and pieces - but none fully.
It's a bit high I think. My exposure (and that has its own sample size issues :) tells me folks are absolutely ignorant about causal inference..esp in industry
September 12, 2025 at 6:20 AM
It's a bit high I think. My exposure (and that has its own sample size issues :) tells me folks are absolutely ignorant about causal inference..esp in industry
1-20% is 57%? Looks a bit unrealistic. Depends on the sample of course.
September 10, 2025 at 8:48 AM
1-20% is 57%? Looks a bit unrealistic. Depends on the sample of course.
Lastly, if you are looking for AI/ML link, the book "Applied Causal Inference Powered by ML and AI" by Chernozhukov et al in 2024 is also good.
August 16, 2025 at 3:02 AM
Lastly, if you are looking for AI/ML link, the book "Applied Causal Inference Powered by ML and AI" by Chernozhukov et al in 2024 is also good.
One of the other texts l like is the "Probability and Causality - Conditional and Average Total Effects" by Rolf Steyer published in 2024.
August 16, 2025 at 2:59 AM
One of the other texts l like is the "Probability and Causality - Conditional and Average Total Effects" by Rolf Steyer published in 2024.
And also "What If.." by Hernan, Robins - keeps updating (latest is in 2025 itself)
August 16, 2025 at 2:55 AM
And also "What If.." by Hernan, Robins - keeps updating (latest is in 2025 itself)
I dont know if you consider it technically as "since 2022", but a new edition came in 2023 - have you seen "A First Course in Causal Inference" by Peng Ding?
August 16, 2025 at 2:54 AM
I dont know if you consider it technically as "since 2022", but a new edition came in 2023 - have you seen "A First Course in Causal Inference" by Peng Ding?
This book has been the best investment I have made in recent years. Just the beginning. I was not aware that Euclid's Elements and mathematical logic had such influence on the works of Locke, Voltaire, Kant, Jefferson, Franklin, Lincoln etc.
May 31, 2025 at 1:07 PM
This book has been the best investment I have made in recent years. Just the beginning. I was not aware that Euclid's Elements and mathematical logic had such influence on the works of Locke, Voltaire, Kant, Jefferson, Franklin, Lincoln etc.
I had never heard about it either !
May 25, 2025 at 7:56 AM
I had never heard about it either !