Your writings on futurism are doubtless a big part of your high ranking (unsurprisingly, LLMs seem to be very interested in the thinkers who wrote about topics like living in a simulation or AI timelines), but a major factor you're neglecting is *publication medium*.
You've produced extremely disproportionately short (= guaranteed fit in context window for training), image-free, easily scraped (at least before you moved to Substack), well-hyperlinked, with a distinctive prose style, while not doing so much in the form of books or paywalled papers or offline non-textual forms of influence, especially video-based. (I am at 723 myself, despite not having published books, being involved in national controversies, etc, largely on the strength of online non-paywalled text like my website and comments, and I expect that my ranking will go up in the future as more recent, larger LLMs get used in such metrics.) Meanwhile, many of the people you cite tried to publish exclusively in old journals which may not even be digitized yet, network TV, etc. 'The future belongs to those who show up in the training corpus.'
Unfortunately, intheweights.com didn't work for my name, Eric S. Harris. (That's how I normally identify myself, ever since the Columbine shootings.)
Worked slightly better when I expanded my middle initial to my middle name and suffixed ", Libertarian". "Eric Stephen Harris, Libertarian" got a Libertarian Party candidate for public office in a different state, Michigan.
But "Eric Stephen Harris, Missouri Libertarian" got a whopping >590 strength · Top 7%<.
I still think this underrates your influence. I wonder if it’s penalizing errors more than I would. But overall the numbers given here seem well calibrated.
Well, that is high. Thanks for analyzing it and explaining it.
I’m just a college professor and haven’t written a book or achieved any social media popularity, but my ranking was way higher than I would have guessed, 665 (top 5%). That’s just based on journal articles. Which we know almost no one reads, so, kind of odd.
My guess is -- 20 years ago I hyphenated my last name on marriage, although my most highly cited publications are still under my original name (Harris). But, the hyphenated name means I
am distinctive and not easily confused with other people with similar last names.
Ok.... just checking the whole distinct name concept, this time input the following query, since from other responses that one can enter descriptive intheweights:
I'm at 178 as TGGP https://intheweights.com/p/tggp whereas the results for "teageegeepea" are deemed possible halluccinations, some of which are definitely not me as I don't have accounts on those platforms. I had hoped not to have others using the same handle.
Your writings on futurism are doubtless a big part of your high ranking (unsurprisingly, LLMs seem to be very interested in the thinkers who wrote about topics like living in a simulation or AI timelines), but a major factor you're neglecting is *publication medium*.
You've produced extremely disproportionately short (= guaranteed fit in context window for training), image-free, easily scraped (at least before you moved to Substack), well-hyperlinked, with a distinctive prose style, while not doing so much in the form of books or paywalled papers or offline non-textual forms of influence, especially video-based. (I am at 723 myself, despite not having published books, being involved in national controversies, etc, largely on the strength of online non-paywalled text like my website and comments, and I expect that my ranking will go up in the future as more recent, larger LLMs get used in such metrics.) Meanwhile, many of the people you cite tried to publish exclusively in old journals which may not even be digitized yet, network TV, etc. 'The future belongs to those who show up in the training corpus.'
'The future belongs to those who show up in the training corpus.' - A quip worth repeating.
(Oh, I do! Probably too recently to show up in most frontier LLMs, though.)
Gwern Branwen
Pseudonymous AI researcher and writer
863 strength · Top 2%
Maybe there is some variation in the scores it gives.
Emil O. W. Kirkegaard
Danish intelligence researcher
770 strength · Top 3%
Maybe. Double checking now, back on June 18th, I also got 863.
Unfortunately, intheweights.com didn't work for my name, Eric S. Harris. (That's how I normally identify myself, ever since the Columbine shootings.)
Worked slightly better when I expanded my middle initial to my middle name and suffixed ", Libertarian". "Eric Stephen Harris, Libertarian" got a Libertarian Party candidate for public office in a different state, Michigan.
But "Eric Stephen Harris, Missouri Libertarian" got a whopping >590 strength · Top 7%<.
Yipee!
https://intheweights.com/p/eric-stephen-harris
I still think this underrates your influence. I wonder if it’s penalizing errors more than I would. But overall the numbers given here seem well calibrated.
Can you give a list of the top 10 people who have influenced you the most?
Well, that is high. Thanks for analyzing it and explaining it.
I’m just a college professor and haven’t written a book or achieved any social media popularity, but my ranking was way higher than I would have guessed, 665 (top 5%). That’s just based on journal articles. Which we know almost no one reads, so, kind of odd.
My guess is -- 20 years ago I hyphenated my last name on marriage, although my most highly cited publications are still under my original name (Harris). But, the hyphenated name means I
am distinctive and not easily confused with other people with similar last names.
Ok.... just checking the whole distinct name concept, this time input the following query, since from other responses that one can enter descriptive intheweights:
CATHERINE HARRIS, PSYCHOLOGY PROFESSOR AT BU
With this info, I am at 2%! with 817 strength!
https://intheweights.com/p/catherine-l~2e~-harris
ok, I tried my full name with my BU affiliation: I now get this!!
918 STRENGTH · TOP 1%
One infers -- if you are in the weights to begin with -- What matters for your strength size --adding on more information.
https://intheweights.com/p/catherine-caldwell~2d~harris
Surprising Cowen does so poorly
Plausibly because he avoids associating with clear strong positions.
We now live in The Age of Em? :-)
Lucky? That's a weird word choice.
956 validates my reading of your work (along with your good friend’s work—Bryan Caplan!). Don’t discount yourself or your value to other people.
I'm at 178 as TGGP https://intheweights.com/p/tggp whereas the results for "teageegeepea" are deemed possible halluccinations, some of which are definitely not me as I don't have accounts on those platforms. I had hoped not to have others using the same handle.
If I type in "isha yiras hashem at substack " I get 698
Ishayirashashem at substack 371
IshaYirasHashem substack 401
IshaYirasHashem or isha yiras hashem gives me nothing but hallucinations
Maybe you typed in Robin Hanson at substack?
No, just "Robin Hanson"
Then I have no clue but it is interesting how small modifications affect et.
I, too, am lucky in the weights at 695.
https://intheweights.com/p/isha-yiras-hashem
Well deserved