Tom Fehring
@tom.fehri.ng
Data person. Formerly at OpenAI, even more formerly an actuary. Posts (rarely) about data, statistical computing, ML/AI, financial markets, urbanism, and/or cats.
Reposted by Tom Fehring
a “Space Force”, if you will
November 13, 2025 at 2:27 PM
a “Space Force”, if you will
This article has popped up on my feed a few times.
Good overview of the pros and cons of different approaches for including not-at-fault crashes, but the numbers are off-base.
TL;DR: data still indicates Waymo is far safer than human drivers.
philkoopman.substack.com/p/blame-free...
1/
Good overview of the pros and cons of different approaches for including not-at-fault crashes, but the numbers are off-base.
TL;DR: data still indicates Waymo is far safer than human drivers.
philkoopman.substack.com/p/blame-free...
1/
Blame-Free Robotaxi Crashes Are Still Crashes
The tricky logic behind robotaxi crash metrics
philkoopman.substack.com
October 5, 2025 at 9:13 PM
This article has popped up on my feed a few times.
Good overview of the pros and cons of different approaches for including not-at-fault crashes, but the numbers are off-base.
TL;DR: data still indicates Waymo is far safer than human drivers.
philkoopman.substack.com/p/blame-free...
1/
Good overview of the pros and cons of different approaches for including not-at-fault crashes, but the numbers are off-base.
TL;DR: data still indicates Waymo is far safer than human drivers.
philkoopman.substack.com/p/blame-free...
1/
Reposted by Tom Fehring
Being a YIMBY means getting regularly accused of trying to complete the Bay Area’s transformation into a tech bro playground, but tbh a big part of the reason why I want more housing here is to RESIST creeping homogenization. I need rent to get cheaper so I can hang out with more weird artists.
January 15, 2025 at 11:48 PM
Being a YIMBY means getting regularly accused of trying to complete the Bay Area’s transformation into a tech bro playground, but tbh a big part of the reason why I want more housing here is to RESIST creeping homogenization. I need rent to get cheaper so I can hang out with more weird artists.
Reposted by Tom Fehring
Part 2: Why do boosted trees outperform deep learning on tabular data??
@alanjeffares.bsky.social & I suspected that answers to this are obfuscated by the 2 being considered very different algs🤔
Instead we show they are more similar than you’d think — making their diffs smaller but predictive!🧵1/n
@alanjeffares.bsky.social & I suspected that answers to this are obfuscated by the 2 being considered very different algs🤔
Instead we show they are more similar than you’d think — making their diffs smaller but predictive!🧵1/n
November 20, 2024 at 5:02 PM
Part 2: Why do boosted trees outperform deep learning on tabular data??
@alanjeffares.bsky.social & I suspected that answers to this are obfuscated by the 2 being considered very different algs🤔
Instead we show they are more similar than you’d think — making their diffs smaller but predictive!🧵1/n
@alanjeffares.bsky.social & I suspected that answers to this are obfuscated by the 2 being considered very different algs🤔
Instead we show they are more similar than you’d think — making their diffs smaller but predictive!🧵1/n