Dean Shamess
@deanshamess.ca
deanshamess.ca
Ph.D. Candidate at JSGS USask // Science Policy, Science of Science, Science and Power and Populism // econ, policy, polisci… whatever works
Ph.D. Candidate at JSGS USask // Science Policy, Science of Science, Science and Power and Populism // econ, policy, polisci… whatever works
That's really 'reasonable' or 'ethical' or not, but I do think it is probably the case that many / most IRBs in Canada would require that.
January 6, 2026 at 3:55 PM
That's really 'reasonable' or 'ethical' or not, but I do think it is probably the case that many / most IRBs in Canada would require that.
preprint papers to demonstrate something about LLMs -> to contacting the authours of those papers, I'm fairly sure our IRB would require us to disclose, in the consent form, both how those survey responses will be stored and used as well as how the initial data was stored, used etc.
IDK if...
IDK if...
January 6, 2026 at 3:55 PM
preprint papers to demonstrate something about LLMs -> to contacting the authours of those papers, I'm fairly sure our IRB would require us to disclose, in the consent form, both how those survey responses will be stored and used as well as how the initial data was stored, used etc.
IDK if...
IDK if...
I think there is something to be said - whether ethical or not - for the fact that many IRBs in many countries might well agree with some part of this complaint as it relates to data storage and retention.
Agreed 100% re: the complaint about IP. However, once this goes from work done using ...
Agreed 100% re: the complaint about IP. However, once this goes from work done using ...
January 6, 2026 at 3:55 PM
I think there is something to be said - whether ethical or not - for the fact that many IRBs in many countries might well agree with some part of this complaint as it relates to data storage and retention.
Agreed 100% re: the complaint about IP. However, once this goes from work done using ...
Agreed 100% re: the complaint about IP. However, once this goes from work done using ...
Reposted by Dean Shamess
Reposted by Dean Shamess
This is contact juggling. There's maybe a few hundred people in the world for whom contact juggling has at one point been their primary source of income, and most of us know each other. It coalesced as an art from in the 1980s and got its name in 1990, though it pulls from practices far older.
January 6, 2026 at 12:27 AM
This is contact juggling. There's maybe a few hundred people in the world for whom contact juggling has at one point been their primary source of income, and most of us know each other. It coalesced as an art from in the 1980s and got its name in 1990, though it pulls from practices far older.
This is very good and worth a read, wherever you land.
The inclusion of Drake dissing is particularly to my taste.
The inclusion of Drake dissing is particularly to my taste.
I wrote about the endless temptation successful people feel to justify and feel justified in a clearly toxic system. Happy new year.
open.substack.com/pub/rottenan...
open.substack.com/pub/rottenan...
Caveat Vendor
Universities, audience capture, and bullshit; Beowulf, Kendrick Lamar
open.substack.com
January 2, 2026 at 10:46 PM
This is very good and worth a read, wherever you land.
The inclusion of Drake dissing is particularly to my taste.
The inclusion of Drake dissing is particularly to my taste.
Reposted by Dean Shamess
I wrote about the endless temptation successful people feel to justify and feel justified in a clearly toxic system. Happy new year.
open.substack.com/pub/rottenan...
open.substack.com/pub/rottenan...
Caveat Vendor
Universities, audience capture, and bullshit; Beowulf, Kendrick Lamar
open.substack.com
January 2, 2026 at 11:00 AM
I wrote about the endless temptation successful people feel to justify and feel justified in a clearly toxic system. Happy new year.
open.substack.com/pub/rottenan...
open.substack.com/pub/rottenan...
Reposted by Dean Shamess
3/ As the division of labor between universities and firms deepens, policymakers increasingly rely on these programs.
Yet we still know surprisingly little about what participation means for scientists themselves.
Yet we still know surprisingly little about what participation means for scientists themselves.
January 2, 2026 at 1:15 PM
3/ As the division of labor between universities and firms deepens, policymakers increasingly rely on these programs.
Yet we still know surprisingly little about what participation means for scientists themselves.
Yet we still know surprisingly little about what participation means for scientists themselves.
Reposted by Dean Shamess
Today is the last day that you can repost this image
December 31, 2025 at 7:48 AM
Today is the last day that you can repost this image
Reposted by Dean Shamess
FACETS Journal is accepting submissions for its 10-year anniversary collection exploring the opportunities and challenges in research. Submit by March 31, 2026. More here ▶️ https://ow.ly/zAa550XLtu6
Guest editors: Jules Blais & Fanie Pelletier
Guest editors: Jules Blais & Fanie Pelletier
December 26, 2025 at 9:40 PM
FACETS Journal is accepting submissions for its 10-year anniversary collection exploring the opportunities and challenges in research. Submit by March 31, 2026. More here ▶️ https://ow.ly/zAa550XLtu6
Guest editors: Jules Blais & Fanie Pelletier
Guest editors: Jules Blais & Fanie Pelletier
If you care to answer... What was the thing about Twin Peaks you felt skeptical about / what overcame that?
December 23, 2025 at 3:32 PM
If you care to answer... What was the thing about Twin Peaks you felt skeptical about / what overcame that?
Reposted by Dean Shamess
A propos of some stuff, a brief thread on randomized experiments. 👇
We like randomization as it provides us with a powerful way to solve a big problem: identification.
Identification should not - cannot - precede other parts of the research process: theory, conceptualization, operationalization.
We like randomization as it provides us with a powerful way to solve a big problem: identification.
Identification should not - cannot - precede other parts of the research process: theory, conceptualization, operationalization.
December 21, 2025 at 1:50 PM
A propos of some stuff, a brief thread on randomized experiments. 👇
We like randomization as it provides us with a powerful way to solve a big problem: identification.
Identification should not - cannot - precede other parts of the research process: theory, conceptualization, operationalization.
We like randomization as it provides us with a powerful way to solve a big problem: identification.
Identification should not - cannot - precede other parts of the research process: theory, conceptualization, operationalization.
I had to update a tonne of R packages this morning... RIP to me and all of the ways none of my work will work anymore 😪
December 18, 2025 at 3:06 PM
I had to update a tonne of R packages this morning... RIP to me and all of the ways none of my work will work anymore 😪
Reposted by Dean Shamess
Call me anti-innovation or whatever, but I prefer my scientists to think like Bengio:
“As scientists, I believe we have a duty to observe current trends in our fields, warn the public and policymakers if significant risks arise, and contribute to potential solutions.”
thelogic.co/commentary/q...
“As scientists, I believe we have a duty to observe current trends in our fields, warn the public and policymakers if significant risks arise, and contribute to potential solutions.”
thelogic.co/commentary/q...
They say Canada’s AI industry has a doom-and-gloom problem - The Logic
Canada is home to some of the finest AI researchers in the world. Yet some prominent figures say negative views of the technology are holding the country back.
thelogic.co
December 15, 2025 at 12:57 PM
Call me anti-innovation or whatever, but I prefer my scientists to think like Bengio:
“As scientists, I believe we have a duty to observe current trends in our fields, warn the public and policymakers if significant risks arise, and contribute to potential solutions.”
thelogic.co/commentary/q...
“As scientists, I believe we have a duty to observe current trends in our fields, warn the public and policymakers if significant risks arise, and contribute to potential solutions.”
thelogic.co/commentary/q...
Probably something to be said for how, at this stage in my fields, you probably don't have anything actually out the door yet - there's no steady stream of completions to go alongside w/e projects you're working on in the early stages atm.
But, who am I to say....
But, who am I to say....
December 14, 2025 at 3:20 PM
Probably something to be said for how, at this stage in my fields, you probably don't have anything actually out the door yet - there's no steady stream of completions to go alongside w/e projects you're working on in the early stages atm.
But, who am I to say....
But, who am I to say....
Academia has felt like a real slog for the last ~18 months.
In the last 2 though, 4-5 projects (some solo, some w/ co's) have all reared their heads again and sprinted towards the finish-ish (aka submission).
While I know there's a long road post-sub to anything, it feels pretty reinvigorating!
In the last 2 though, 4-5 projects (some solo, some w/ co's) have all reared their heads again and sprinted towards the finish-ish (aka submission).
While I know there's a long road post-sub to anything, it feels pretty reinvigorating!
December 14, 2025 at 3:19 PM
Academia has felt like a real slog for the last ~18 months.
In the last 2 though, 4-5 projects (some solo, some w/ co's) have all reared their heads again and sprinted towards the finish-ish (aka submission).
While I know there's a long road post-sub to anything, it feels pretty reinvigorating!
In the last 2 though, 4-5 projects (some solo, some w/ co's) have all reared their heads again and sprinted towards the finish-ish (aka submission).
While I know there's a long road post-sub to anything, it feels pretty reinvigorating!
Gems like rdr2 and witcher 3. Which, to their own, are not exactly masterpieces of storytelling - but I feel like are good comparators as high polish, open-ish, branching narratives.
December 13, 2025 at 4:28 PM
Gems like rdr2 and witcher 3. Which, to their own, are not exactly masterpieces of storytelling - but I feel like are good comparators as high polish, open-ish, branching narratives.
Agreed! I played a lot of BG3, a LOT, but that was because of the gameplay and experimentation. The best I can say about the story, and this isn’t a backhand, is that it did constantly feel like I was being propelled.
But, as far as actual storytelling, it fell so far short of other recent(ish)…
But, as far as actual storytelling, it fell so far short of other recent(ish)…
December 13, 2025 at 4:28 PM
Agreed! I played a lot of BG3, a LOT, but that was because of the gameplay and experimentation. The best I can say about the story, and this isn’t a backhand, is that it did constantly feel like I was being propelled.
But, as far as actual storytelling, it fell so far short of other recent(ish)…
But, as far as actual storytelling, it fell so far short of other recent(ish)…
Happily, and with great anticipation every day!
Zero sugar monster (the green one, not the white) runs laps around Australian-brewed flat whites without breaking a sweat.
Zero sugar monster (the green one, not the white) runs laps around Australian-brewed flat whites without breaking a sweat.
December 11, 2025 at 7:22 PM
Happily, and with great anticipation every day!
Zero sugar monster (the green one, not the white) runs laps around Australian-brewed flat whites without breaking a sweat.
Zero sugar monster (the green one, not the white) runs laps around Australian-brewed flat whites without breaking a sweat.
Fuck, this is so good.. Good for Harald 😂
December 7, 2025 at 7:12 PM
Fuck, this is so good.. Good for Harald 😂
Reposted by Dean Shamess
Political polarization on confidence in the scientific community accelerated in the 2024 General Social Survey, with moderates now showing a significant downturn.
(Apologies if I missed this updated with 2024 data already)
/1
(Apologies if I missed this updated with 2024 data already)
/1
December 6, 2025 at 4:24 PM
Political polarization on confidence in the scientific community accelerated in the 2024 General Social Survey, with moderates now showing a significant downturn.
(Apologies if I missed this updated with 2024 data already)
/1
(Apologies if I missed this updated with 2024 data already)
/1
And their belief will be correct
December 5, 2025 at 2:41 PM
And their belief will be correct
benefit of this? Or at least some sort of .... humility in saying "we understand so little about the world so if we can make ID claims, we can more humbly assert that we only really know this one small piece of some much bigger thing"?
December 2, 2025 at 11:13 PM
benefit of this? Or at least some sort of .... humility in saying "we understand so little about the world so if we can make ID claims, we can more humbly assert that we only really know this one small piece of some much bigger thing"?
evidence about the social world that's obviously not good.
Describing the assumptions of identification (should be) quite simple. Describing the assumptions of non-identified quant work inherently introduces many more caveats that we're trained to poke holes in. So, perhaps, there is also some ...
Describing the assumptions of identification (should be) quite simple. Describing the assumptions of non-identified quant work inherently introduces many more caveats that we're trained to poke holes in. So, perhaps, there is also some ...
December 2, 2025 at 11:13 PM
evidence about the social world that's obviously not good.
Describing the assumptions of identification (should be) quite simple. Describing the assumptions of non-identified quant work inherently introduces many more caveats that we're trained to poke holes in. So, perhaps, there is also some ...
Describing the assumptions of identification (should be) quite simple. Describing the assumptions of non-identified quant work inherently introduces many more caveats that we're trained to poke holes in. So, perhaps, there is also some ...