Ehud Karavani (@ehudkar) profile picture

Machine Learning ๐Ÿค–๐ŸŽ“. Causal Inference ๐Ÿ”ดโ†’๐ŸŸ โ†๐ŸŸก. Creator of ๐™ฒ๐šŠ๐šž๐šœ๐šŠ๐š•๐š•๐š’๐š‹. Was active in a pseudonym account, Now at https://teapawt.net/t.co/ZJDLboXmUn

Joined July 2016
Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)Photo by Ehud Karavani (@ehudkar)
None can match my matching slide ๐Ÿ˜…๐Ÿ‘จโ€๐ŸŽ“ #causalinference #epitwitter
1
16
2
62
Ehud Karavani retweeted
Prediction is not causation. @ehudkar explains all about it.
1
1
89
I'll be speaking tomorrow at @PyData TLV about causal inference - providing a conceptual overview and briefly touring causallib. Last few tickets remaining. grab yours at pydata.org/telaviv2024
3
106
Knowledge comes from falsifying hypotheses, constraining on what cannot be true (and limiting the flexibility of imagination). Yet deep learning engineers think just pouring more data "and letting the model figure it out" gonna solve it. Are we surprised LLMs hallucinate?
1
88
What an excellent opportunity to revise my answer and introduce some more nuance to it. First, admittedly my prior is often hype-averse so it is fair to say I'm a LLM skeptic at the moment, and I don't believe they currently hold much *scientific* value.
This Post is from an account that no longer exists. Learn more
1
1
81
LLMs might be inefficient query engines, but so is asking Siri; natural language itself is quite inefficient - but it's natural. And conversing with the Internet is a convenient way for humans to surf the Internet.
1
34
To conclude, LLMs' scientific value is over-hyped but that are still useful if you understand their limitations and think critically. Namely, they are a new, natural, and convenient interface to query the Internet, which will only improve as factuality will be improved.
29
I'm humbled by Alex asking me to interview and thankful for him persisting (I had to cancel our first try due to child-induced Covid ๐Ÿ˜…). Super excited about this Monday!
This Post is from an account that no longer exists. Learn more
1
4
689
Everyone is very excited for their new bedtime story ๐Ÿ™ƒ with its colorful drawings and excellent content, all by the wonderful @AleksanderMolak
1
612
haha, using ChatGPT to bootstrap an initial causal graph to later be critiqued and refined by an expert is just the manifestation of Cunningham's Law: "The best way to get the right answer on the Internet is not to ask a question; it's to post the wrong answer"
This Post is from an account that no longer exists. Learn more
1
95
With all due respect to Frank, this is the equivalent of demanding a large survey of disinterested experts to compile a list of priors for a Bayesian analysis. No, instead we ask for a handful of experts and do prior predictive checks and sensitivity analyses. >>
Replying to @ehudkar @statsepi
Pick 10 OSs at random from respected journals. Iโ€™ll bet not one did the necessary step of surveying a large number of disinterested experts to compile the list of potential confounders, then including all of those in prospective data collection.
1
143
I cannot disagree with an assessment saying it is easier to produce a garbage observational study than a garbage RCT. But this is exactly why target trials should be encouraged as a framework for observational studies. teapawt.net/Lach_cribb/status/1786โ€ฆ
2
113
All in all, I think that's a weak criticism of observational studies. And that there are better ones out there. A personal favorite is this one: teapawt.net/ADAlthousePhD/status/1โ€ฆ
...and sometimes those eligibility requirements which make representativeness difficult in RCT's are in fact quite critical to ensuring validity (in a causal sense) of the result - representing a narrower target population than "everyone" to be sure - but we do this for a reason!
96