Daniel Jones
@dcjones.bsky.social
Computational biologist at Fred Hutchinson Cancer Center (See my work here: https://dcjones.github.io/). I also post about baking, books, and Seattle politics.
"We often blame bureaucrats or red tape for the country’s paralysis. But the fault lies less in the rules than in who has the resources to weaponize them."
open.substack.com/pub/data4dem...
open.substack.com/pub/data4dem...
How Regulation by Litigation Strangled American Abundance
Environmentalists aren't blocking housing. Wealthy homeowners are.
open.substack.com
December 20, 2025 at 4:53 PM
"We often blame bureaucrats or red tape for the country’s paralysis. But the fault lies less in the rules than in who has the resources to weaponize them."
open.substack.com/pub/data4dem...
open.substack.com/pub/data4dem...
I went all out optimizing proseg this week, and after many rounds of profiling and tinkering, the new release is as much as 1.5-4x faster and reduces memory usage by around 10%.
More importantly, I discovered and fixed a subtle bug that reduced segmentation accuracy for large gene panels!
More importantly, I discovered and fixed a subtle bug that reduced segmentation accuracy for large gene panels!
December 19, 2025 at 6:03 PM
I went all out optimizing proseg this week, and after many rounds of profiling and tinkering, the new release is as much as 1.5-4x faster and reduces memory usage by around 10%.
More importantly, I discovered and fixed a subtle bug that reduced segmentation accuracy for large gene panels!
More importantly, I discovered and fixed a subtle bug that reduced segmentation accuracy for large gene panels!
Reposted by Daniel Jones
tfw the guy who made this chart is going to be in charge of the fed
December 11, 2025 at 4:11 PM
tfw the guy who made this chart is going to be in charge of the fed
Feel like pure shit just want her back
en.wikipedia.org/wiki/Good_Ne...
en.wikipedia.org/wiki/Good_Ne...
Good Neighbor policy - Wikipedia
en.wikipedia.org
December 10, 2025 at 7:20 PM
Feel like pure shit just want her back
en.wikipedia.org/wiki/Good_Ne...
en.wikipedia.org/wiki/Good_Ne...
I'm somehow just discovering that there's pretty compelling deep learning framework being built in rust burn.dev/books/burn/m...
Why Burn? - The Burn Book 🔥
burn.dev
December 7, 2025 at 5:51 PM
I'm somehow just discovering that there's pretty compelling deep learning framework being built in rust burn.dev/books/burn/m...
Reposted by Daniel Jones
Congratulations to Seattle's Sound Transit, America's fastest-growing train network, which is opening an 8-mile, 3-station extension to Federal Way today!
This is the system's 4th new segment opened in less than 2 years, continually pushing ridership up to new record highs!
This is the system's 4th new segment opened in less than 2 years, continually pushing ridership up to new record highs!
December 6, 2025 at 8:56 PM
Congratulations to Seattle's Sound Transit, America's fastest-growing train network, which is opening an 8-mile, 3-station extension to Federal Way today!
This is the system's 4th new segment opened in less than 2 years, continually pushing ridership up to new record highs!
This is the system's 4th new segment opened in less than 2 years, continually pushing ridership up to new record highs!
Reposted by Daniel Jones
Pew's main findings:
- Newer multifamily is safer than older multifamily
- Newer multifamily is safer than newer single-family
- Single-stair multifamily buildings are as safe as buildings with two stairwells
www.lewis.ucla.edu/2025/12/03/1...
- Newer multifamily is safer than older multifamily
- Newer multifamily is safer than newer single-family
- Single-stair multifamily buildings are as safe as buildings with two stairwells
www.lewis.ucla.edu/2025/12/03/1...
Episode 103: Fire Safety in Multifamily Housing with Alex Horowitz (Incentives Series pt. 6)
Explore why multifamily homes are often safer from fires than single-family housing, as Alex Horowitz shares research on building age, fire risk, and single-stair safety in our Incentives Series.
www.lewis.ucla.edu
December 3, 2025 at 5:44 PM
Pew's main findings:
- Newer multifamily is safer than older multifamily
- Newer multifamily is safer than newer single-family
- Single-stair multifamily buildings are as safe as buildings with two stairwells
www.lewis.ucla.edu/2025/12/03/1...
- Newer multifamily is safer than older multifamily
- Newer multifamily is safer than newer single-family
- Single-stair multifamily buildings are as safe as buildings with two stairwells
www.lewis.ucla.edu/2025/12/03/1...
I'm never going to be able to afford memory again. Might just pivot all my work towards probabilistic data structures and online stats.
www.reuters.com/business/mic...
www.reuters.com/business/mic...
Micron to exit consumer memory business amid global supply shortage
Memory chipmaker Micron Technology said on Wednesday it will exit its consumer business, as it doubles down on advanced memory chips used in artificial intelligence data centers amid a global supply s...
www.reuters.com
December 3, 2025 at 7:26 PM
I'm never going to be able to afford memory again. Might just pivot all my work towards probabilistic data structures and online stats.
www.reuters.com/business/mic...
www.reuters.com/business/mic...
Lingonberry cruffins. Going full Scandinavia mode lately.
November 17, 2025 at 5:05 PM
Lingonberry cruffins. Going full Scandinavia mode lately.
Reposted by Daniel Jones
Seen our Transit Tracker project on social media? You can learn how to build your own here: transit-tracker.eastsideurbanism.org
Transit Tracker
Transit Tracker is a DIY customizable public transit arrivals board for your home.
transit-tracker.eastsideurbanism.org
November 16, 2025 at 5:46 AM
Seen our Transit Tracker project on social media? You can learn how to build your own here: transit-tracker.eastsideurbanism.org
Got my copy of the inaugural issue of Seattle-centric baking magazine Crumb. It's good! www.readcrumb.co
November 15, 2025 at 7:43 PM
Got my copy of the inaugural issue of Seattle-centric baking magazine Crumb. It's good! www.readcrumb.co
Reposted by Daniel Jones
The last five months with Claude Code have completely changed how we work.
matsen.group/agentic.html details:
• How agents work (& why it matters)
• Git Flow with agents
• Using agents for science
• The human-agent interface
Questions? What has your experience been?
matsen.group/agentic.html details:
• How agents work (& why it matters)
• Git Flow with agents
• Using agents for science
• The human-agent interface
Questions? What has your experience been?
Agentic Coding For Scientists
A four-part series on using coding agents like Claude Code for scientific programming, covering fundamentals, workflows, best practices, and the human side of AI-assisted development.
matsen.group
November 13, 2025 at 5:08 PM
The last five months with Claude Code have completely changed how we work.
matsen.group/agentic.html details:
• How agents work (& why it matters)
• Git Flow with agents
• Using agents for science
• The human-agent interface
Questions? What has your experience been?
matsen.group/agentic.html details:
• How agents work (& why it matters)
• Git Flow with agents
• Using agents for science
• The human-agent interface
Questions? What has your experience been?
Reposted by Daniel Jones
seattle's political geography is as wild as (and presumably at some level is reflective of) its physical geopraphy
November 13, 2025 at 12:37 AM
seattle's political geography is as wild as (and presumably at some level is reflective of) its physical geopraphy
Cosign this. I was frustrated on election night when no seemed to be able to say what exactly the range likely outcomes was.
Fortunately @elections.kingcounty.gov provides a lot of historical data. I had fun digging into it with @ronpdavis.bsky.social and learning it's actually fairly predictable!
Fortunately @elections.kingcounty.gov provides a lot of historical data. I had fun digging into it with @ronpdavis.bsky.social and learning it's actually fairly predictable!
WAY too many of our political pundits wrote as if history barely exists--acknowledging the "progressive shift" but saying little to nothing about how big it usually is, and acting as if Katie was super behind and maybe-just-barely had a chance.
Those narratives were almost all mirage.
Those narratives were almost all mirage.
November 12, 2025 at 3:50 AM
Cosign this. I was frustrated on election night when no seemed to be able to say what exactly the range likely outcomes was.
Fortunately @elections.kingcounty.gov provides a lot of historical data. I had fun digging into it with @ronpdavis.bsky.social and learning it's actually fairly predictable!
Fortunately @elections.kingcounty.gov provides a lot of historical data. I had fun digging into it with @ronpdavis.bsky.social and learning it's actually fairly predictable!
Reposted by Daniel Jones
Like many of us, I've been waiting with baited breath for the Mayoral results for a week.
Putting together the Wilson PAC & raising the majority of its funds made me even more invested & amped up the roller coaster.
Despite my emotional journey, the result was predictable, (though not certain).
Putting together the Wilson PAC & raising the majority of its funds made me even more invested & amped up the roller coaster.
Despite my emotional journey, the result was predictable, (though not certain).
November 12, 2025 at 3:11 AM
Like many of us, I've been waiting with baited breath for the Mayoral results for a week.
Putting together the Wilson PAC & raising the majority of its funds made me even more invested & amped up the roller coaster.
Despite my emotional journey, the result was predictable, (though not certain).
Putting together the Wilson PAC & raising the majority of its funds made me even more invested & amped up the roller coaster.
Despite my emotional journey, the result was predictable, (though not certain).
Really wasn't a sure thing a few drops ago (it could have followed one of these shallower historical trajectories and missed the mark), but it's certainly nearly a sure thing now.
November 12, 2025 at 12:32 AM
Really wasn't a sure thing a few drops ago (it could have followed one of these shallower historical trajectories and missed the mark), but it's certainly nearly a sure thing now.
Reposted by Daniel Jones
Katie Wilson leads by 1346 votes!
November 12, 2025 at 12:05 AM
Katie Wilson leads by 1346 votes!
Updating my plot on Katie Wilson's latest numbers compared to previous races. Progressive vote share sometimes drops a little in the last few percent, but it's extremely unlikely that it's enough to reverse her lead at this point.
November 11, 2025 at 12:56 AM
Updating my plot on Katie Wilson's latest numbers compared to previous races. Progressive vote share sometimes drops a little in the last few percent, but it's extremely unlikely that it's enough to reverse her lead at this point.
They're calling him the Nate Silver of Seattle.
(Except I learned stats and epistemology from books, not internet poker, and I haven't cooked my brain with twitter addiction.)
(Except I learned stats and epistemology from books, not internet poker, and I haven't cooked my brain with twitter addiction.)
Last big drop of ballots hits this afternoon, so we'll see if I'm right, but I think Katie Wilson is going to win (by a narrow margin).
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
here's my attempt, with the red line showing where wilson would end up if she wins the same portion of remaining ballots that she won in the latest batch — it ends up being just a 0.2% margin
November 11, 2025 at 12:03 AM
They're calling him the Nate Silver of Seattle.
(Except I learned stats and epistemology from books, not internet poker, and I haven't cooked my brain with twitter addiction.)
(Except I learned stats and epistemology from books, not internet poker, and I haven't cooked my brain with twitter addiction.)
Reposted by Daniel Jones
Incredible. Progressive challenger Katie Wilson now has a 91-vote lead in Seattle’s race for mayor. She had been steadily gaining on incumbent Bruce Harrell.
There are about 8,000 ballots left to count.
There are about 8,000 ballots left to count.
November 10, 2025 at 11:57 PM
Incredible. Progressive challenger Katie Wilson now has a 91-vote lead in Seattle’s race for mayor. She had been steadily gaining on incumbent Bruce Harrell.
There are about 8,000 ballots left to count.
There are about 8,000 ballots left to count.
Reposted by Daniel Jones
Last big drop of ballots hits this afternoon, so we'll see if I'm right, but I think Katie Wilson is going to win (by a narrow margin).
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
here's my attempt, with the red line showing where wilson would end up if she wins the same portion of remaining ballots that she won in the latest batch — it ends up being just a 0.2% margin
November 10, 2025 at 5:05 PM
Last big drop of ballots hits this afternoon, so we'll see if I'm right, but I think Katie Wilson is going to win (by a narrow margin).
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
Turns out the ongoing debate about the dangers of increased surveillance in the city was not entirely hypothetical.
There was no sign ICE accessed Redmond's Flock system before Monday's arrests, but city officials said residents remained concerned about the technology.
Redmond turns off Flock Safety cameras after ICE arrests
www.seattletimes.com
November 10, 2025 at 7:38 PM
Turns out the ongoing debate about the dangers of increased surveillance in the city was not entirely hypothetical.
Last big drop of ballots hits this afternoon, so we'll see if I'm right, but I think Katie Wilson is going to win (by a narrow margin).
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
here's my attempt, with the red line showing where wilson would end up if she wins the same portion of remaining ballots that she won in the latest batch — it ends up being just a 0.2% margin
November 10, 2025 at 5:05 PM
Last big drop of ballots hits this afternoon, so we'll see if I'm right, but I think Katie Wilson is going to win (by a narrow margin).
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
As shown here, she needs to as well or better in the last 17% of the ballots as she did on Friday's drop. The good news is that this is pretty likely.
Continuing to look at the late leftward swing in Seattle elections. Here's Katie Wilson's trajectory compared to every off-year election since 2017 (where there was a clear left candidate). Historically the swing usually surpasses what she needs to win (black line).
November 7, 2025 at 11:52 PM
Continuing to look at the late leftward swing in Seattle elections. Here's Katie Wilson's trajectory compared to every off-year election since 2017 (where there was a clear left candidate). Historically the swing usually surpasses what she needs to win (black line).
Reposted by Daniel Jones
Thanks to @dcjones.bsky.social, I have key info!
The progressive percentage of the Friday vote count is typically much higher on Friday than Thursday.
Moon's was 9% higher on Friday than Thursday.
Gonzales' was 5%
Wilson primary was 14%.
Thursday to Monday they were 12%, 4.5%, and 12%.
The progressive percentage of the Friday vote count is typically much higher on Friday than Thursday.
Moon's was 9% higher on Friday than Thursday.
Gonzales' was 5%
Wilson primary was 14%.
Thursday to Monday they were 12%, 4.5%, and 12%.
November 7, 2025 at 2:38 AM
Thanks to @dcjones.bsky.social, I have key info!
The progressive percentage of the Friday vote count is typically much higher on Friday than Thursday.
Moon's was 9% higher on Friday than Thursday.
Gonzales' was 5%
Wilson primary was 14%.
Thursday to Monday they were 12%, 4.5%, and 12%.
The progressive percentage of the Friday vote count is typically much higher on Friday than Thursday.
Moon's was 9% higher on Friday than Thursday.
Gonzales' was 5%
Wilson primary was 14%.
Thursday to Monday they were 12%, 4.5%, and 12%.