@peternka
 Oxford, OH
peternencka.com/
Joined December 2008
  • Tweets29
  • Following2.4K
  • Followers1.1K
9Media
Haiku 5.5 is very very good at handwriting recognition. It is now the frontier inexpensive option for our historical handwriting recognition benchmark opencensus.us/check/standing…
8
326
Peter Nencka 📊 retweeted
Ryan Hill@RyanReedHillSep 28
Imagine you're a kid growing up in 1910. You happen to live next door to a doctor. Are you more likely to become a doctor by 1940 than your friend Carl, who lives 5 doors down? Our new paper suggests that yes, by about 30%!
15
96
19
1,003
80,274
Apply! Tell your friends! Have them also tell their friends!
Jim Flynn@jim_flynn9Sep 28
We are hiring 2 tenure-track applied micro AP roles @Miami_Econ! One in health One more general applied micro Come join a dynamic department in a beautiful college town, less than an hour from Cincinnati.
2
10
1,086
Analogous to the partnership with refine.ink, journals should team up with pangram and provide an AI overview for all submissions and referee reports

Refine — AI Verification Trusted by World-Class Experts

Good decisions require verified quality. Refine devotes hours of frontier compute to protect your work and reputation from fixable mistakes.

refine.ink
3
240
At the Economic History Association meetings this weekend in St Louis if anyone wants to chat about this! (or anything really)
Peter Nencka 📊@peternkaSep 17
🚨 Announcing OpenCensus, a new project with @EzraKarger We are transcribing the historical US censuses and releasing them as open data 2.6 million people so far in 1940, browse here: opencensus.us
6
982
aside: it's interesting that Google has basically the whole frontier for handwriting transcription in our application
Peter Nencka 📊@peternkaSep 17
Replying to @peternka
Here's how the models are doing so far relative to their cost, both on the overall sample and in "hard" fields where context is key
3
767
🚨 Announcing OpenCensus, a new project with @EzraKarger We are transcribing the historical US censuses and releasing them as open data 2.6 million people so far in 1940, browse here: opencensus.us
9
121
6
453
40,959
more replies
Thanks also to our all-star list of contributors and advisors. We are trying to make this the broadest possible team If you want to help, let us know! opencensus.us/get-involved
1
5
626
PS -- this is not a replacement for IPUMS! Their data is amazing, cleaned up, and available for all years. We are focusing on a pipeline to create open data that can be used by everyone to improve and expand existing datasets
1
10
529
Load more