Michael Littman
@mlittmancs
Followers
8K
Following
3K
Media
94
Statuses
1K
Joined November 2011
It makes me happy that the common machine-learning phrase "test set" is a palindrome.
20
245
2K
Airline installs chatbot. Customer gets bad information from it. Customer asks for refund. Airline says "the chatbot is a separate legal entity that is responsible for its own actions" (!). Court says no. (Whew.) Airline makes refund and turns off chatbot.
18
198
927
Nice blog post (co-authored by a Brown PhD student!) explaining how systems like ChatGPT are built using clever combinations of unsupervised (for language), supervised (for preferences), and reinforcement (for applying the latter to the former) learning:
5
91
420
Charles's's @NeurIPSConf keynote from last night can be viewed even if you aren't registered for the conference. Don't let the title fool you---it's a sociotechnical romp. Actually, don't let the presenter name fool you either---there's more there's, too.
2
90
407
Filled out my passport renewal and was surprised that three slots---hair color, eye color, and employer---all had the same answer.
3
6
372
My book, "Code to Joy: Why Everyone Should Learn a Little Programming", has been sent to MIT Press for production. Woo! It covers elements of programs, how to use them to solve problems, and how ML/AI can help make programming easier for all of us. Should be out Fall 2023!.
14
26
310
@sydgibs My wife named our first child Max. I begged her to let me name the 2nd one Min. I'm afraid you might be right.
9
12
302
A paper I reviewed yesterday asserted that the quality of a researcher's reviews declines over the years. At least, I think that's what it said. I didn't read it that closely because I was pretty sure I'd end up rejecting it just like 100s and 100s of others I've read before.
6
11
314
Is it just me, or have weekdays become the time to have meetings with the purpose of gathering action items to work on over the weekend? It seems wrong. I need a better plan.
14
12
300
Happy to be celebrating my 2^5 wedding anniversary today. The power of two people celebrating a power of two!.
9
2
301
Starting a new "rotating" position shortly as Division Director of Information and Intelligent Systems at NSF (AI, HCI, ML, etc.). I'll be learning about what the research community needs, what NSF provides, and how to bring the two closer together!.
18
12
284
I had an idea for a book, wrote a proposal, negotiated with publishers, signed a contract, wrote a draft, worked with an editor, got peer reviewed, arranged for image rights, edited the index, folded in copyeditting feedback, waited a little longer, and POOF, it's real! Magic!
15
17
275
I wrote a book! I met today with my editor at MIT Press and she says she thinks it's ready to be sent out for review, 3 months ahead of schedule. Woo! If all goes smoothly, it'll be on sale Fall 2023. (She said books, like "human babies" take about 9 months to gestate.).
5
7
267
Tonight, I attended the ceremony for the Turing Award (the "CS Nobel"). It was a great honor to see @ylecun , @geoffreyhinton , and Yoshua Bengio recognized for their stunning contributions (and leadership). After, I entertained Yann by spinning my new Fellow's certificate.
2
18
253
A former undergrad (now a CS PhD student at a top school) gave me, maybe?, a very nice complement. Roughly: Everyone thought your machine-learning class was too easy, but I realize now that I actually learned a lot and retained it!.
4
2
225
Email sent to me to work with a Stanford or MIT prof on an AI project "to enhance your college application profile this fall!" I guess it's never too late for me to stress out about getting into a good college.
3
2
204
Like other fields, Machine Learning has some serious concerns about replicability. A concrete step is to nudge the culture to one where reproducing important results is a norm. I'm excited that my new PhD student, Jessica Forde, is organizing an event to do JUST THAT!.
🎉We are pleased to announce the 3rd edition of the reproducibility challenge at #NeurIPS2019! 🎉. Researchers and students at all levels are encouraged to participate. Registration Deadline: Nov. 1 2019.Report Submission Deadline: Dec. 1 2019.
4
37
193
I wasn't sure if I was allowed to mention it, but MIT Press made an announcement, so I guess it's ok (!). I'm under contract to write a broad-audience book on computing: Why we all need to learn to code and how AI will make programming a computer more like teaching a person.
We're excited to showcase just a few of our recent book signings, slated for publication in 2022 and beyond: #newbooks #TBR
20
9
193
Introducing a new video series with me and @isbellHFh . We're watching Westworld together and commenting on the AI/CS/ML ideas they touch on. We call the series "Smoov and Curly on Video"! Here's the teaser: . First episode drops tomorrow.
12
21
199
Brown named me a 2020 Distinguished Research Achievement Award winner. Wow, thanks! @BrownHCRI @BrownBigAI.
Brown Daily Herald @the_herald writes about the three @BrownUniversity professors who won 2020 Research Achievement Awards in Distinguished category: @BrownCSDept Michael Littman @mlittmancs, @HealthyBrown @BrownCAAS Peter Monti, @BrownMedicine John Sedivy
13
5
197
Nice summary of this year's NeurIPS conference: . Yes, that's me in the silly hat. It's for science. @BrownHCRI.
3
31
178
We did it! The 2021 AI100 Report is out! My deepest appreciation to my coauthors, who did their fantastic research and writing on top of all the stresses and constraints of a global pandemic. I am in awe of their accomplishments and brilliance!.
5
38
173
My colleagues and I studied recent advances in AI for this year's #AI100Report, to be released 9/16! Join me and some of the report authors and organizers to discuss key findings at either of two sessions on 9/28. Registration is free!
2
40
171
Went to campus today but forgot to bring a mask and forgot my office key. But Plan B is working out well.
6
1
160
I'm academically expecting! I've long said that one of my biggest goals is to have a grandstudent. Well, I learned today that my former PhD student @vukosi is currently advising a PhD student. I'll share the ultrasound whenI get it. :-) @BrownHCRI.
3
1
155
I got to carry a torch at Waterfire in Providence on behalf of RLDM 2022 (Reinforcement Learning and Decision Making). Note the subtle reference to the cartpole problem, a standard benchmark in the field.
5
6
149
Discovered today that the dog has learned to recognize when I end a Zoom call. She uses this skill to know when it's ok to ask for a walk.
3
1
145
A glowing review of my book (@mitpress) from one of my science communication heroes! (I went to one of @stevenstrogatz 's book events on the notion of infinity and he encouraged me to write a book, and I did!).
I’ve just started reading this book, and it’s fantastic! Super informative and funny. I don’t know the author, Michael Littman, personally, but it’s obvious from every page that he is a marvelous teacher. HIGHLY Recommended!
3
8
128
@isbellHFh and I are thrilled that we were both named as ACM Fellows today. a great honor in the CS field. (We nominated each other. Because we're friends and we look out for each other.) Thought our #omscs @GTOMSCSCS peeps might appreciate the news. :-)
21
12
126
So proud of my friend Charles Isbell @isbellHFh , who will be U. Wisconsin's next provost, starting this Fall. For those not steeped in academic terminology, that's basically the school's CEO (one step below president).
8
5
124
Speaker at a talk just said that researchers who don't engage with the current work on large scale neural models will be dinosaurs. I'm wondering if I get to pick which kind of dinosaur before I have to decide.
11
4
108
Just had my first week as an NSF division director. I started keeping track of initialisms (like "NSF" for "National Science Foundation") and acronyms (like "DAD" for Deputy Assistant Director) that people use. Didn't take long to hit 90. This is going to be a wild ride!.
5
1
113
My chair has a name now! @BrownHCRI.
Congratulations to the latest @BrownCSDept recipients of named chairs! @mlittmancs is now the Royce Family Professor of Teaching Excellence in Computer Science and Ellie Pavlick is the Manning Assistant Professor of Computer Science:
7
6
110
Here's my first computer science parody song in a few years. I hope you like it!
6
16
105
In studio recording an introductory machine learning course for The Great Courses. @BrownBigAI @BrownHCRI
3
5
100
I saw a college friend yesterday and we were reminiscing about the old days. Her: "I had forgotten that you were a mime.". Me: "That's ok. I don't talk about it much.".
4
1
101
I'm working on my office for the first time in 1 1/3 years. I am having . a lot of feelings.
1
0
95
If anyone is curious, it looks like my talk at NeurIPS (Deep RL Workshop) will be livestreamed: Saturday (today, 12/14) 8pm East Coast time. @BrownHCRI.
1
10
93
Thurs was a big day for me at the NSF: (1) Worked with DARPA on a joint effort. (2) Spoke to Italy at a science cooperation mtg. (3) Attended the Dept. of Commerce roll out of NIST's AI risk management framework. (4) Rehearsed for a presentation to be held in the US Congress.
1
1
90
When asked "how likely are you to recommend <company> to a friend or colleague?", I always say "not at all likely". After all, the probability that anyone would ask my opinion on the topic within a relevant window is essentially zero. Perhaps I'm answering too literally?.
5
1
78
Yoshua Bengio gets it right---we've seen progress in AI, there's a long way to go, we should be working to ensure positive societal implications, we should work together, and we should be thinking about learning causal structure! @BrownHCRI.
1
22
80
I just found out that my book, Code to Joy, has a release date! It should be available October 3rd. Woo! I've been working with MIT Press on the index, the endorsements, and finalizing the fonts in the pseudocode examples. When the cover design is finished, I'll post it.
3
8
82
RLDM is coming back! Please send us your reinforcement learning papers and come join us in Providence this summer!.
Excited to announce the #RLDM2022 call for workshops: with workshop co-chair @michaelbrowni19. RLDM is one of my favorite conferences: focused on reinforcement learning and decision making, highly interdisciplinary - short: a great venue for your workshop!.
0
15
77
Another big AI announcement from the White House. This one includes the AI R&D strategic plan update, which I co-chaired!
3
20
80
So excited to be hosting RLDM @ Brown!.
This week, @BrownUniversity hosts the Multi-disciplinary Conference on Reinforcement Learning and Decision Making (RLDM), bringing more than 500 thought leaders to PVD and demonstrating Brown’s continued status as a leading center for RL research and study
1
8
75
A favorite moment from #IJCAI2022 . Stuart Russell guided me and Bart Selman and Carla Gomes to a statue in honor of AlphaGo!
1
2
74
Here's an award-receiving selfie. (Definitely not an award-winning selfie!) #IJCAI22 @BrownBigAI
2
3
79
My PhD student @dabelcs successfully defended virtually his dissertation on abstraction in reinforcement learning! At Brown, students get a rubber chicken upon completion, so I transmitted a virtual one to him on Zoom. @BrownHCRI @BrownBigAI.
1
2
70
I am very honored to be recognized and to get to present my work to the university this Spring! Thanks, Chris Paxson @BrownUniversity ! @BrownHCRI @BrownUResearch.
Out of more than 700 faculty members, two are chosen each year for Brown's Presidential Faculty Award, and this semester, President Christina Paxson selected Professor @mlittmancs of Brown CS. Michael is the first @BrownCSDept CS recipient of the award:
3
3
72
Milestone reached. Last Spring, I began writing scripts for a professionally produced introduction to ML online course. I just submitted the 1st draft of the concluding lecture. A lot to do before production begins, but glad to have gotten to this stage. @BrownHCRI @BrownBigAI.
4
2
69
Spending the summer lecturing online was hard work, but I do appreciate this tribute from one of my students. (The carrots are a nod to a running joke in lecture as students frequently responded to my questions with "^" in the Zoom chat.)
2
2
66
Not that you should judge a book by its over, but mine is pretty nice. :-) Official release date is October 3rd, available for pre-order at Amazon, Barnes & Noble, Indigo, and other sites: !
1
12
70
I mean, technically, it was the 80s. But, I don't think I'd been called OG before, so that's nice!.
This Thursday at the #DRLworkshop . we are happy to host Michael Littman @mlittmancs.He is one of the OGs of reinforcement learning. He's been publishing and researching #RL since the 1990s and taught #MOOCs on the subject
5
6
67
Maybe one reason people fear AI is that CS hasn't had its reckoning yet. Chemistry had mustard gas, physics had the atom bomb, biology had eugenics. It's time for us CS professionals to act like people whose work and teaching impact society. @BrownHCRI.
1
20
60
I've joined up with 11 other AI experts as well as staff at @aaas to engage in a year-long exploration of public engagement in AI. I hope to have lots to share in the coming months!. @BrownHCRI @BrownBigAI.
4
12
62
Controversial proposal to require all CS authors to include negative broader impact statements in their papers and proposal. Good to see the computing community grappling with these issues---we won't get it right if we don't try. @BrownHCRI @ACM_FCA.
2
29
59
To get re-certified for the human-in-the-loop ML work I'm doing, I had to read the Belmont Report (1979), which lays out the ethical issues of studying people. I was struck by how beautifully written / reasonable / broadly relevant it is! @BrownHCRI.
3
7
62
Science asked me to comment on DeepMind's latest game playing AI---Capture the Flag in Quake. @BrownHCRI.
1
16
59
I'm so lucky to know @isbellHFh and his truly wonderful family. He paints a clear and heartbreaking picture of what FAR too many people are going through: "It really doesn’t have to be the way it’s always been." @BrownHCRI @BrownBigAI.
0
6
57
My book, Code to Joy, came out last week and is currently #15 on Amazon's list of books on AI!
3
5
56
Getting ready to speak at IJCAI 22. Apparently, when you get to Austria, they revert the spelling of your last name back to how it looked when your family left several generations ago. #IJCAI22
2
2
54
With democracy struggling worldwide, I stand with my brilliant former PhD student Kavosh Asadi who is spreading the word about protests happening in Iran.
2
11
54
So honored to get a university-level teaching award from Brown!
Brown has recognized our faculty with two Excellence in Teaching Awards: Professor @mlittmancs won the Philip J. Bray Award for Excellence in Teaching in the Physical Sciences, and Paul Valiant won the Dean's Award for Excellence in Teaching.
4
1
55
On a call with some UK AI colleagues. Apparently, they name everything there after Turing and it's starting to get confusing (Turing Institute, Turing Fellowship, etc.). But I don't think I knew one of the coolest ones. Turing is on money!
1
2
54
The median decision problem---is the given element of a duplicate-free list the median of the list?---can be solved in linear time. But, for even-length lists, the complement problem is trivially O(1) for all possible elements. After all, every one is a co-median.
2
6
54
Proud of colleague George Konidaris, co-winner of the IJCAI-JAIR best paper prize. His work makes progress on a core AI concern---bridging between muddy perceptual inputs and symbolic representations that support decision making. @BrownBigAI @BrownHCRI.
1
4
54
What's an example of an ordinary problem that you were able to solve using programming/computation?.
27
2
53
I'm co-chairing RLDM with the inimitable @cate_hartley . If you're into reinforcement learning (or Montreal), I hope you can submit an abstract and/or attend the conference. @BrownHCRI.
I'm excited to announce that the 4th multi-disciplinary Reinforcement Learning and Decision Making conference (#RLDM2019) will be in Montreal this July 7-10! Check out the speaker list, and consider submitting an abstract (deadline March 1st)! Visit: 🧠💻
1
17
53
My thoughts on what statistics can tell us about being human, delivered roughly a year ago in the form of a commencement speech: .
5
4
53
Nice article about ethical considerations in AI research, as well as how @NeurIPSconf tried to engage with it. What we're seeing "is not a crisis in the public understanding of science, but a crisis in science’s understanding of the public”. @BrownHCRI.
0
9
49
Super excited to be chosen as an AAAS Leshner Fellow. I will spend a chunk of time this coming year working on AI Science Communication! @BrownHCRI @BrownBigAI @BrownCSDept.
It is our honor to introduce the newest cohort of @aaas #LeshnerFellows, focusing on #ArtificialIntelligence!. The incoming fellows work in a diverse range of disciplines & are ready to discuss how #AI affects our lives every single day!. #SciEngage.
1
3
51
Is there an established correlation between (pandemic-induced) isolation and increases in feelings of self loathing? Asking for DEFINITELY not-a-friend.
10
1
44
Oh! It looks like my Introduction to Machine Learning for The Great Courses is out of post-production and has been announced! @BrownHCRI.
1
5
47
I loved "The Ethical Algorithm" (audiobook). Clear prose, important topic, informative examples. Unlike many books on cutting edge topics, this one has just gotten more and more relevant since it came out (in November). @mkearnsupenn @Aaroth @BrownHCRI.
1
12
50
Finished Week 2 at NSF. Focused on "DD concur"---the process of going through the funding recommendations of the program directors and signing off on them as Division Director: $88M in requests, $17M approved (!). I'm in awe of the great proposals and the brilliant reviewers.
0
1
46
I’m just getting over COVID. The last few days, I was angry and I couldn’t understand why anyone was talking about anything other than the pandemic. How is this still a problem?? Some people lose their sense of taste, some their sense of smell. I lost my sense of proportion.
4
1
46
@isbellHFh and I won the AI educator(s) of the year award for our joint work! Our award talk at EAAI/AAAI is now posted, including a ten-minute collection of clips from our joint videos and a Q&A:
5
4
47
If you or someone you know wants to learn some basics about machine learning and computer science, check out my new public talk! (two sessions via Zoom Tue 9/8 and 9/15 @ 12pm PT / 3pm ET). I'd love to see you there! Tickets: . @BrownBigAI @BrownHCRI.
0
14
47
Deep learning could improve their robot hugs by 23.7% or more! (Actually, we once used a machine-learning approach to detect whether a person was interested in a high five, fist bump, or hug from the robot and had it respond appropriately.) @BrownHCRI.
0
20
43
Big White House announcement about AI came out today. I wrote some text for the "fact sheet", but it got substantially changed before release. Still, it was fascinating to watch this process unfold!
1
8
40
When I got to Brown in 2012, CS was not one of the ten most popular concentrations. Now, ten years on, one in 5 students graduate with a CS degree. It really is a fast-moving field. .
@BrownCSDept just gave diplomas to 366 undergrads, 21.8% of all the undergrads who graduated from @BrownUniversity this May. It's a record high number of graduations from a single department or school at Brown and for the 6th year, CS is the most popular concentration on campus.
1
2
41
NYT: "Apple is worth $3 trillion—more than Walmart, Disney, Netflix, Nike, Exxon Mobil, Coca-Cola, McDonald’s, AT&T, Goldman Sachs, Boeing, IBM and Ford COMBINED." Wha? I'm struggling to wrap my head around how that could be. Close on its heals is Microsoft and other tech giants.
4
2
39
Yes! We made it across the twenty twenty twenty twenty one boundary! Even just saying it aloud feels like a real accomplishment. .
0
0
38
Nicely written article about image synthesis with an emphasis on its potentially dangerous role in society. Second time I've seen the idea that synthesis may result in self surveillance. @BrownHCRI.
1
17
37
Kinda cool. I haven't seen the book in the wild yet, but apparently it's even on display in some places!.
0
0
36
This article on Scrabble isn't about AI or reinforcement learning. But, it is a lovely example of how hard it is to design rewards so that they produce the desired incentives. If there's one thing everyone ought to know about RL, it's that. @BrownHCRI.
0
6
38
I've said it before and I'll say it again, no one likes redundancy.
1
0
36
I'm definitely in favor of kids studying music! (That said, I think the title is misleading AND misses the point of why coding is important---we should all be able to express to machines what we want them to do, it's empowering).@BrownHCRI.
5
2
37
Proud to work with the talented Evan Cater, selected for the Randy Pausch Summer Research Award. From my rec: Evan is at the stage I want my 3rd year PhD students to reach, in spite of being a 3rd year undergrad. Amazing! @BrownHCRI.
2
1
38
Charles Isbell @isbellHFh is the 2023 Tapia Award Winner for Scientific Scholarship, Civic Science, and Diversifying Computing! . With his recent Nico Haberman award and becoming Provost at the U. Wisc., it's been a very good year for him professionally!.
1
1
36
TIL, one of the earliest formal treatments of the logic of time---reasoning about what comes before what---was undertaken in the 1950s. The researcher's name: Arthur Prior.
2
1
35