Kevin Dorst
@kevin_dorst
Followers
4,605
Following
729
Media
171
Statuses
2,271
Philosopher at @MIT , trying to convince people that their opponents are more reasonable than they think. Blog:
Cambridge, MA
Joined January 2020
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
PistangPinoyKorea with SB19
• 120864 Tweets
ロックの日
• 86545 Tweets
エプソムC
• 38777 Tweets
#RIZIN47
• 32496 Tweets
#ユーフォ3期
• 31855 Tweets
#ModiCabinet
• 27120 Tweets
카리나 시구
• 22220 Tweets
エプソムカップ
• 19882 Tweets
レーベンスティール
• 17994 Tweets
プロ初勝利
• 14889 Tweets
#njDOMINION
• 13369 Tweets
Pinned Tweet
"Rational Polarization" is now forthcoming in The Philosophical Review!
(Shorter version: )
5 years in the making; still lots more to figure out. Here's the TLDR, and a detailed summary:
37
153
744
I'll be joining the MIT philosophy department in the fall! Super excited
42
11
646
New post: "Ideological Bayesians"
TLDR: It’s often said that Bayesian updating is unbiased and converges to the truth—and, therefore, that biases must emerge from non-Bayesian sources. That’s wrong. For agents like us, updating on our *total* evidence...
10
114
488
Whittling updates (exercising my new-found gift superpower):
10
7
206
Do we teach epistemology wrong?
When we lead with skepticism, the arc is: "PROBLEM. Then [a bunch of complicated stuff to get us back to square one]." Intro epistemology as recovery.
But there's SO MUCH epistemology that starts where students are and goes to NEW places...
17
7
196
"Rational Polarization" is now officially published in The Philosophical Review!
Here's a thread on how rational Bayesians can fall into predictable polarization when evidence is ambiguous:
"Rational Polarization" is now forthcoming in The Philosophical Review!
(Shorter version: )
5 years in the making; still lots more to figure out. Here's the TLDR, and a detailed summary:
37
153
744
2
26
165
New post: "In Defense of Epistemic Empathy". I argue against standard reasons given for thinking your political opponents are irrational.
Are there any I missed?
7
40
155
I’m launching a new blog on politics, polarization, psychology, and (ir)rationality. The first post argues that demonization may be more of a problem than polarization, and that irrationalist narratives in psychology could be to blame:
9
38
150
I'm starting a new blog series called "Reasonably Polarized: Why politics is more rational than you think."
TLDR: We can't coherently blame polarization on irrationality, so we need a new story. I'm going to try to tell it over the coming weeks.
This...
21
25
134
New Post: "Bayesians Commit the Gambler's Fallacy".
Upshot: Rational people who start out with causal uncertainty and update on unbiased data will (on avg) expect random processes to "switch" more than they actually do—just like real people...
9
17
126
New post!
Like humans, ChatGPT commits the conjunction fallacy, the gambler's fallacy, a form of confirmation bias ("belief inertia"), and exhibits apparent overconfidence. What to make of it?
12
26
121
Fake hands and unmarked clocks—turns out IKEA is a better supplier of epistemological props than you’d think.
1
2
114
I've finally posted a full draft of my "Rational Polarization" manuscript:
Feedback welcome!
TLDR: §1: Me and my childhood friends were predictably polarized. I think this process was broadly rational.
§2: Most theories of rational polarization fail...
4
31
113
Correlation doesn't imply causation.
But causation implies correlation. (Other things equal.)
And things are evidence for the things that imply them. (Bayesianism.)
So correlation is evidence for causation. Right?
24
11
106
Seriously, someone once read a footnote, thought "That was too easy" and then BAM:
Endnotes.
1
4
103
New Stranger Apologies post:
"Centrists are (probably) less biased"
I am not a centrist! Still:
TLDR: Some empirical studies suggest the left is more responsive to political evidence; others suggest the center. Yet some very general dynamics suggest...
6
22
100
...Polarization and the import of cognitive biases. Disagreement and irrelevant influences. Ideology and social distortion. Standpoint theory and epistemic injustice. Network epistemology and the dynamics of social trust.
We could have intro epistemology as discovery.
So...
1
0
92
"Being Rational and Being Wrong" is forthcoming in Philosophers' Imprint!
TLDR: Empirical studies often find that people are "miscalibrated"—eg of all the claims they're 80%-confident in, only 60% are true. This is taken as evidence for OVERCONFIDENCE....
3
10
91
Why not a philosophical-stylist course? One where we read great stylists of different types (Lewis; Nozick; Quine...[suggestions?]), and then students have to practice writing short papers like them. Try on different stylistic hats; find your voice
27
2
85
Just came across this great piece on what "confirmation bias" really is---and whether it's actually so pervasive and problematic---by
@jesswhittles
, which tells the story of her PhD on the topic and how she changed her mind. Definitely worth a read!
4
20
80
Just posted the syllabi and handouts from the courses I taught last semester:
- Intro epistemology ()
- Seminar on human (ir)rationality ()
The epistemology course was a bit of an experiment: rather than starting with skepticism...
5
8
76
New post: "The Natural Selection of Bad Vibes (Part 1)"
TLDR:
Things seem bad. But chart-wielding optimists keep telling us that things good. What gives?
Hypothesis: the point of conversation is to *solve problems*, so public discourse will focus...
5
22
76
Happy to announce that Pitt Philosophy is organizing a summer program for students from underrepresented groups, to help get a sense of grad school in philosophy:
Open Doors Philosophy Academy
Fully funded! App deadline March 15.
Please spread the word!
2
47
74
Watch out. 1.5 months of paperwork, “ethics” tests, and thumb-twiddling, and I finally have IRB approval to ask people questions like “which city do you think is bigger—Denver or San Francisco”.
The power is intoxicating.
2
4
69
I'm inclined to think grad students should worry a bit less about which *problems they find interesting*, and a bit more about which *methods they find tractable*.
Important work is the product of interesting-Qs + productive-methods. In philosophy we have TONS of interesting...
3
5
68
TLDR. Real-world polarization is *predictable*: we can often predict how our actions will influence our own opinions.
We have no theory of how this could be rational.
I prove that rational Bayesians can predictably polarize iff evidence is AMBIGUOUS.
That allows them to be...
2
7
66
Imagine if students were as excited about probabilism as they are about skepticism...
6
3
63
Stranger Apologies is back! Now on Substack.
TLDR: We’ve lost our epistemic empathy. Irrationalist narratives from psychology are partly to blame. Recent work from psychology and philosophy changes the narrative. Sign up to see why.
5
16
57
Sometimes I find myself wondering “is Bayesianism *really* all that great?”
Then I talk to non-Bayesians.
7
7
60
...I'm thinking about dropping skepticism and the analysis of knowledge (almost?) entirely from my course next time.
Thoughts?
Especially curious to hear the reasons folks think this could be a bad idea!
20
0
60
Continuing my polarization-lit deep dive, this paper by
@cailinmeister
and James Weatherall is fantastic:
.
Fact: people's political beliefs are correlated, even when they have nothing to do with each other. (Pro-lifers are usually pro-gun.)
Explanation:
5
13
55
"Good Guesses", with
@MMandelkern
, is now forthcoming in PPR!
TLDR:
1) We guess all the time;
2) We do it by optimizing a tradeoff between accuracy and informativity;
3) This yields new theories of belief, assertion, and the *conjunction fallacy*...
2
15
55
When I was a kid, if I stared at a clock I as always surprised by how slow the second-hand seemed to move. Now if I do, it feels like it’s flying.
Is this was getting old feels like?? (But seriously, any studies on the phenomenology of time by age?)
11
5
55
"Assertion is Weak", with
@MMandelkern
, is now forthcoming in Philosophers' Imprint!
TLDR: Lots of people think the norm of assertion is strong—you should only assert P if you know it, are justified in being sure of it, etc. We argue they're wrong...
6
10
54
...*asymmetrically* sensitive to the truth—more able to recognize evidence on one side than the other.
I prove that this can lead to predictable, persistent polarization.
Then I use an experiment and simulations argue that this mechanism helps explain real-world polarization.
1
6
54
Just published a piece on
@ArcDigi
, making the argument that we shouldn't explain polarization by appeal to irrationality.
TLDR: You can't think think your own (polarized) political beliefs are irrational; but we know that both sides form their...
8
12
54
Does simply owning something make you value it more?
This apparent *endowment effect* violates standard econ models, and was one of the main findings that led to the rise of behavioral econ.
But have the findings been misinterpreted? A thread 🧵
5
21
51
"Which I will demonstrate by showing how easy it is to offer a conservative argument in < 280 characters."
Why is Twitter so much more Left-wing that the population at large? Here's a theory: it lends itself to angry, self-righteous and emotive statements. Conservative arguments are generally too nuanced to fit into 280 characters.
904
468
4K
5
3
47
Dear philosopher friends,
I write, with regret, to inform you that I am moving on. You see, I just calculated my first confidence interval on *real world DATA*. With that, I have leveled up to the rank of (social) scientist, and must now think myself better than you.
Yours,
K
4
1
49
I'm no metaphysician, but I love this new "Making Metaphysics" paper by Thomas Byrne in Phil Imprint:
TLDR: Breaking something ≠ causing it to break; killing someone ≠ causing them to die; etc. Instead they involve "making"—a relation which, it turns...
2
12
48
Does it bother you as much as me that these nutrition facts are inconsistent?
7
2
47
People often say formal epistemology is too idealized. Why? Economics, game theory, cognitive science, and psychology often use equally idealized (say, Bayesian) models to study how *real* people think. Hypothesis: people confuse two notions of “idealization”. (1/2)
6
6
46
New paper coming out in Analysis!
"(Almost) All Evidence is Higher-Order Evidence" (w/ Brian Hedden):
TLDR: Higher-order evidence is evidence about what’s rational to think in light of your evidence. Many have argued that it’s special—falling into its...
1
9
47
Need literature help!
Planning a social-epistemology course on *bias*, and I'm looking for good papers (accessible to non-technical folks) on
(1) what bias is, and
(2) how to understand its effects in social-network settings (Algorithmic-fairness stuff welcome!)
Suggestions?
20
6
45
Every now and then I start wondering if I’m a logician.
Then I read work by logicians.
1
1
43
Happy to announce that it's now been 14 MONTHS since we (
@Chris_Dorst
) submitted a manuscript to Thought, 6 months since they started saying "initial review any day now", and 2 months since they said "no, seriously".
Maybe if we leave it there we can get some sort of record.
6
0
43
That feeling when you come back to a draft after months and realize you can cut almost 20% of the words and lose only 2% of the content...
"How did past-Kevin think this sentence was necessary??"
1
1
42
Read Piketty’s Capital, and found myself pretty convinced by it all. Curious to hear criticisms!
Socialists: why think his proposed solutions (global tax on capital, etc) aren’t radical enough?
Capitalists: why think they’re too radical?
Any further reading refs most welcome
12
3
41
This week's Reasonably Polarized post: "Confirmation Bias Maximizes Expected Accuracy". ()
TLDR: Why do people interpret information in a way that favors their prior beliefs? Because interpreting evidence requires searching it for flaws, and there's a...
3
12
37
Knowledge-first or accuracy-first epistemology?
The weather apps are taking sides:
1
2
40
Now, in more detail:
§2:
Obvious fact: real-world polarization is predictable. Example: you know that if you read only biased sources, this will bias your opinions.
Surprising fact: no extant rational theory allows this.
Standard Bayesian theories don't. They imply a...
1
1
39
Finally got the chance to read
@RobertTalisse
's "Overdoing Democracy."
It's a good book! Quick summary, and a couple thoughts:
The cool political-philosophy move is the point that insofar as we need some version of *deliberative* democracy for political legitimacy... (1/9)
2
6
39
Just posted a substantially revised version of my paper on overconfidence. New title: "Being Rational and Being Wrong" ().
TLDR: even on the strongest theories of rationality, the connection between being rational and being right is looser than...
3
11
39
Kahneman was brilliant, and an inspiration. Even those of us who spend our time critiquing his ideas have learned so much from him.
Go read the Tversky and Kahneman classics—they're a delight.
It's rare for a scientist to so profoundly influence their own field, and even rarer to do so in another field.
Kahneman, who passed away at 90, didn't merely criticise economics from the outside for its flawed understanding of human behaviour. He engaged with it, and
4
82
330
1
5
38
Okay okay, this whole Wordle thing is getting a bit too close for comfort. Now it’s got word-completions, polarization, AND confirmation bias. Basically covers the whole “rational polarization” project I’ve been going on about, w/o having to read a 60-page paper…
This whole "the NYT ruined Wordle!" thing is a great case study in confirmation bias. It also gives us some insight into human nature that helps explain how conspiracy theories work.
1K
13K
98K
3
4
36
I was always on board in principle, but it took
@fitelson
years of pestering me before I finally took the time to learn how to code. Now, most of the projects I’m working on wouldn’t be possible without it. Best investment of research time I’ve made
It's not uncommon to hear people say that computer simulations "just aren't philosophy."
No more!
Conor Mayo-Wilson and I have a new paper where we argue that computer simulation should be seen as a *core* philosophical method.
16
92
383
2
2
37
This week's Reasonably Polarized post: "How to Polarize Rational People"
TLDR: I report on experiment showing how ambiguous evidence can be used to polarize people—and claim that I can even polarize YOU, my (rational) readers.
When it comes to...
3
10
37
That’s the plan!
(And thank you :).)
Lmao
@kevin_dorst
's work on polarisation is so good wtf. Formal social epistemology will save the world.
3
2
48
0
0
34
New running route—not bad! :)
Starting to think Pgh doesn’t get enough credit for how pretty a city it is
4
0
36
Found myself returning to this wisdom from
@lastpositivist
recently. Like so many academics, I've recently wanted to write on things relevant to our current crises (and felt hopeless when I couldn't). That was a mistake, I think. And maybe one relevant to others. (1/7)
It’d be good for Discourse if people separated more clearly stakes of an issue from stakes of us discussing it. Like, for pretty much any important political issue how society behaves affects wellbeing of millions. But you and I disagreeing right now aren’t deciding for society.
4
10
95
2
2
35
Nice summary of some recent work on (ir)rationality:
The reason wars are over. Reason won by Hanno Sauer
0
10
34
New paper!
The Conjunction Fallacy: Confirmation or Relevance?
with WooJin Chung,
@MMandelkern
, & Salvador Mascarenhas
TLDR: 2 exps suggest that conversational relevance drives the conjunction fallacy. This supports the guessing-based theories.
Longer:
1
8
33
Come to the Formal Epistemology Workshop (FEW2023) in Boston June 12–14! It'll be at Northeastern University, co-hosted by MIT.
We have a fantastic lineup of talks. See the schedule and register to attend here:
Please spread the word!
0
11
33
...is impossible—instead, we must *selectively attend* to certain questions, ignoring others. Yet correlations between what we see and what questions we ask—“ideological” Bayesian updating—can lead to predictable biases and polarization.
Why can't we use our total evidence?
1
2
32
"Deference Done Better", with
@ben_levinstein
, Bernhard Salow (),
@brookehus
, and
@fitelson
is now forthcoming in Phil Perspectives!
TLDR: There are lots of "experts" we should (epistemically) defer to—rational opinions, chances...
1
8
31
I wish I could agree! But the small talk / informal run-ins seem to be impossible to replicate. That might seem minor, but it’s one of the core reasons for grad students to attend conferences—it’s how they get their names out there and begin to feel part of the discipline. So...
I confess. I would never have believed how good an online conference could be if I had not been forced to try it out. Now inclined to repeat it even if I don’t have to.
11
5
140
7
2
31
As an ex-gamer, this analysis seems spot on.
About 10 years ago I found myself transposing the gaming mindset to academia (building a CV is like leveling-up a character). Still haven’t worked out how to feel about the change... (Do you discuss,
@add_hawk
?) Will have to read!
My book, GAMES: AGENCY IS ART is out! It's about:
How game designers sculpt agency.
How games let us record, transmit, and explore new forms of agency.
How real games make us more free.
How gamification undermines our freedom.
64
255
2K
2
1
30
"Myside bias" is the tendency to generate more arguments for the side of a debate you believe than for the side you don't ()
Why is this a bias?
Shouldn't we expect those who believe P to be those who know more arguments favoring P than disfavoring it?
10
6
29
On HS quality: has anyone figured out why we have to wait till college to learn about validity, soundness and arguments in premise/conclusion form? They can teach us multivariate calculus but “are there any cases where the Ps are true but C false?” is too high-level??
1
3
29
Does it bother you as much as me that "probable" and "probably not" (and "almost certain" and "almost certainly not") are not symmetric around 50%?
(From )
1
3
30
Excited to announce a new horrible variant on Polish notation.
In the step from online-only version to published version, Analysis decided to swap every multiplication sign (•) for an ampersand (&).
Suspect this'll be my big break.
3
1
30
Upshot: we have to *selectively attend* to some questions, and ignore others that are equally available to us.
Epistemologists in critical tradition have been telling us this forever—that ideology works, in part, by shaping our patterns of attention
We can use this insight...
1
3
30
Just had my first in-person meeting with a student in... a long time. Forgot how much better it is to talk philosophy in person than over zoom...
0
0
27
Just posted an updated version of the polarization paper!
Significantly shorter, includes a 1-page sketch of how it works at the beginning, and removes (some! Not all. Never all) of the stylistic infelicities.
More feedback always welcome!
I've finally posted a full draft of my "Rational Polarization" manuscript:
Feedback welcome!
TLDR: §1: Me and my childhood friends were predictably polarized. I think this process was broadly rational.
§2: Most theories of rational polarization fail...
4
31
113
1
3
28
Agreed—I really think reviewers should focus more on whether there’s insight behind a paper than whether it’s water-tight (and that editors should encourage that!)
“There is something bizarre about these acceptance rates. Eric Schliesser has observed that “this low rate is only defensible if you think that publication in philosophy has the kind of risk where any false positive leads to society’s catastrophe”
2
6
42
5
0
27
New post: The Natural Selection of Bad Vibes (Part 2: Social Media)
TLDR: The vibes are bad, even though—on most ways we can measure—things are (comparatively) good. The last post showed how disproportionately-negative sharing can emerge from trying...
5
4
28
Achievement unlocked: a "Dorst and Dorst" paper now exists! Just posted a new brotherly draft with
@Chris_Dorst
, arguing that fine-tuning is evidence for design. (He convinced me of atheism 15 years ago... Who'd have thought we'd end up writing this one?)
2
4
27
...'Reflection' (Martingale) principle: your estimate of your future opinion must match your current opinion (You should "price in" the bias of your sources).
Extant theories that deny Reflection use updates that you expect to decrease accuracy—so, arguably, aren't rational.
1
1
27
...empirical findings: people are most inclined to be biased when evidence is mixed, complex, or hard to interpret—exactly the situations where evidence is plausibly ambiguous.
Can we make good on this connection?
2
2
26
Of course, this is still highly theoretical.
But I've (predictably!) convinced myself, at least, that the core hypothesis—that ambiguity-asymmetries can drive polarization—has enough going for it that is deserves a closer empirical look.
I'm hoping to start that work...
2
4
26
2024 FEW (formal epistemology workshop) website and submissions are now live!
0
6
26
Mike Titelbaum in a seminar just now: "Some people are too quick to let ordinary reasoners off the hook".
!!
Like a knife to my heart.
3
2
26
PSA: your friendly reminder that people with common priors CAN agree to disagree. (Aumann's theorem has been misinterpreted.)
@LedermanHarvey
3
3
27
When, after waiting for >1 month, a journal emails you saying "We haven't yet received your proof corrections and they are now overdue; please submit them within one business day."
This is the first email they've sent me about it.
2
1
27
Can't not tweet this. How can you not love philosophy? "Retweeting: its linguistic and epistemic value":
2
8
27
Just learned Mind has a 1-submission-per-year policy (like Phil Review, similar to PPR and Nous's submission blackout).
I'm inclined to think this is dumb—if they have too many submissions, they should publish more articles, and maybe even(!) incentivize reviewers.
Thoughts?
5
0
26
The madman strategy for R&Rs:
Write an incredibly gracious, substantive, and long (as long as the paper) cover letter explaining in (exhausting) detail how seriously you've taken each comment.
Subtext: "If you don't recommend accept, you may have to read another such letter..."
1
2
26
Just uploaded a handbook article () on higher-order evidence for a Routledge volume (edited by
@cmlittlejohn
and Maria Lasonen-Aarnio!). Tries to summarize a lot of the recent Bayesian-friendly work on the issue and say...
1
2
25
My sense is there are a lot of philosophers writing good public-facing philosophy on how to think about the pandemic. What are some of your favorites that we should all check out / promote? I'll start: O'Connor and Weatherall on misinformation about Covid
2
6
26
Fascinating new paper coming out in Mind on the connection between decision theory and desire! (h/t Milo Phillips-Brown)
0
6
26
Interesting post! Why not instead an optimistic take on the trends?:
Analytic philosophy is unified by trying to make intractable issues tractable. As a result (1) we shouldn't expect progress in many areas at once, and (2) the applied turn is evidence that there is a class...
I wrote up a blog post with my totally invented history and sociology of analytic philosophy. I am a pessimist about my own field, and I don't think my own work has been any better or in any way escapes the trends that make me something of a downer.
74
87
583
1
1
25