Max Langenkamp Profile Banner
Max Langenkamp Profile
Max Langenkamp

@mslkmp

Followers
929
Following
575
Media
180
Statuses
724

biosecurity, cryptography, meaning, tofu

Boston, MA
Joined April 2021
Don't wanna be here? Send us removal request.
Explore trending content on Musk Viewer
Pinned Tweet
@mslkmp
Max Langenkamp
2 years
Tapestry (thread of threads)
1
2
7
@mslkmp
Max Langenkamp
8 months
this would be funnier if it weren't so common
Tweet media one
28
181
2K
@mslkmp
Max Langenkamp
6 months
Terence Tao on how he solves math problems
Tweet media one
8
87
843
@mslkmp
Max Langenkamp
5 months
still contemplating this
Tweet media one
24
104
755
@mslkmp
Max Langenkamp
8 months
from @meaningness "statistics and the replication crisis". I can attest to having been both the scientist begging for the right test and the witness of someone told to just "get the stats to work"
2
3
159
@mslkmp
Max Langenkamp
2 years
I present to you all: "How Open Source Machine Software Shapes AI" (essay/my thesis/ @AIESConf paper - links at end). In it, @dhadfieldmenell , @dyuephoria , and I claim that Machine Learning OSS (MLOSS) is a crucial but neglected factor shaping AI development. 1/
Tweet media one
3
11
75
@mslkmp
Max Langenkamp
2 years
After blistering hours of bibtex debugging, I have submitted my Master's thesis! There is still much to do before it's ready for public consumption, but wanted to share + hint at what I've been fostering for the last 18 months. 1/3
Tweet media one
9
3
66
@mslkmp
Max Langenkamp
11 months
can confirm Helen is one of the most compassionate clear sighted people I've ever known. I don't have strong feelings about the conflict but it has been interesting to see the ire heaped upon her
@QiaochuYuan
QC
11 months
no. none of you assholes understand helen toner in the slightest. she's one of the most thoughtful and principled and level-headed people i met in the bay. not sucked into the crazy psychodrama that i could tell, not motivated by money or status. she's a homie
42
12
541
2
1
49
@mslkmp
Max Langenkamp
6 months
For me the key here is how different Tao's description is from the dry stereotype I was taught of 'how math works'. Terence's description here feels rich and of the world!
2
0
47
@mslkmp
Max Langenkamp
2 years
There's a lot of excitement about AI and automation. Mostly, people seem to assume *if* a task can be automated and there's economic incentive, it will be. Here's a short essay on why I think that perspective is naive.
Tweet media one
2
3
40
@mslkmp
Max Langenkamp
2 years
One common quip I hear among technologists is that 'AGI/BCI/etc is inevitable'. This claim misunderstands the contingency of tech development and is attempting to be a monopoly in the marketplace of ideas. 1/
Tweet media one
4
8
39
@mslkmp
Max Langenkamp
1 year
“Collecting art is one of the most important things you can do” somehow this throwaway sentence from @tylercowen has stuck with me in the last two years and immensely improved my appreciation of art
6
4
38
@mslkmp
Max Langenkamp
1 year
some thoughts on wearing clothes as practice
Tweet media one
4
0
33
@mslkmp
Max Langenkamp
10 months
Did some crackpot physics last night — I think it's pretty plausible that pigeons detect the earth's magnetic field by detecting deflection of Ca ions by flapping their wings
Tweet media one
4
1
32
@mslkmp
Max Langenkamp
1 year
After recently graduating, @lukeigel and I were reflecting on the role of MIT on the world at large. We decided to record a dialogue that included our personal experience, and also what Luke learned while making the @mitregressions documentary.
Tweet media one
1
2
27
@mslkmp
Max Langenkamp
2 years
I spent hundreds of hours in the last two years exploring different therapeutic modalities to try and understand/address the psychological pain I was in...
2
2
22
@mslkmp
Max Langenkamp
2 years
5 years and 2 pieces of paper later, I did it! At this rate I'll have 8 degrees by 40 when I can finally retire (1/commentary)
Tweet media one
1
0
20
@mslkmp
Max Langenkamp
3 years
Noticing that a large fraction of the great thinkers on impact of technology were Christian. MacLuhan (Catholic convert), Illich (Catholic priest) and Ellul (Christian anarchist) are obvious examples.
4
2
21
@mslkmp
Max Langenkamp
2 years
Two years ago, @tuomasoi and I took a class on decision theory. Recently, we went on a hike and had a dialogue about decision theory.
Tweet media one
1
1
20
@mslkmp
Max Langenkamp
2 years
Truly bizarre scene earlier this month at the Senate hearing on risks and opportunities from AI. It began with a Senator (Sen. Johnson) explaining that the Republicans are boycotting the hearing because … Jordan Peterson wasn’t allowed to call in remotely to testify on AI.
Tweet media one
2
1
18
@mslkmp
Max Langenkamp
2 years
I rest at night a little easier knowing that people like @j_asminewang @saffronhuang and @divyasiddarth are working on tech governance. Thrilled and hopeful for this project
@collect_intel
Collective Intelligence Project
2 years
Today, we are publishing the Collective Intelligence Project whitepaper, our R&D agenda for transformative tech governance. It is our contribution to what we see as one of the world’s biggest problems. We couldn’t be more excited to share it with you. 1/
11
126
403
2
0
18
@mslkmp
Max Langenkamp
6 months
Happy spring! Hope you get to go outside and feel the sunlight fill your eyes
Tweet media one
0
0
18
@mslkmp
Max Langenkamp
1 year
On this #FourthofJuly , I have come from the technological frontiers bearing a gift, a surprise, an item of commerce. As an ode to the founding fathers, the American spirit, and Lady Liberty, I present to you: 🇺🇸PATRIOT PROTEINS
Tweet media one
1
0
15
@mslkmp
Max Langenkamp
2 years
During my qualitative interview of a burner, PLUR came up. V interested in sub cultural meaning-making like this
Tweet media one
1
1
15
@mslkmp
Max Langenkamp
2 years
oh god my thesis is due today
0
0
14
@mslkmp
Max Langenkamp
3 years
Wow it really feels like spring in Boston and I can’t stop smiling
Tweet media one
1
0
12
@mslkmp
Max Langenkamp
2 years
Loving @meaningness ’s latest book on AI risk. I’ve been thinking about this since 2017 and the book clarified a lot of confusion I had, particularly around how we conceive of AI (though I have some objections too). Thread: 1/
@Meaningness
David Chapman
2 years
🆕 🤖 📕 How to avert an AI apocalypse... and create a future we would like. A book about mostly-overlooked AI risks and how you can help avert them. The full text is now on the web!
Tweet media one
32
65
336
1
3
13
@mslkmp
Max Langenkamp
7 months
how to estimate the speed of light using cheese and a microwave: 1/
Tweet media one
1
2
14
@mslkmp
Max Langenkamp
3 years
all models are wrong but, oh Father, who art in heaven, hallowed be thy Name, please let my model be right
0
1
12
@mslkmp
Max Langenkamp
1 year
Astounded that IFP managed to set up this partnership with NSF. Might be the most promising development in metascience in the last decade
@calebwatney
Caleb Watney
1 year
I could not be more excited about @IFP 's new partnership with @NSF to help them design and execute experiments on improving scientific grantmaking. NSF has shown tremendous leadership in this area, and we are thrilled to support their efforts.
Tweet media one
14
37
209
0
0
13
@mslkmp
Max Langenkamp
2 years
Here's the central diagram from my essay. It is approximately how I think of different therapeutic modalities
Tweet media one
2
2
12
@mslkmp
Max Langenkamp
2 years
Last note is that I'm going to be presenting a version of my thesis at AI Ethics and Society @AIESConf in Oxford Aug 1-3. If you're near Oxford + interested do reach out 3/3
0
1
13
@mslkmp
Max Langenkamp
6 months
Tu Youyou, who won a Nobel prize for her characterization of antimalarial compounds, had her breakthrough in trying to replicate ancient Chinese alchemical texts
Tweet media one
1
0
12
@mslkmp
Max Langenkamp
3 years
Can’t wait to lie next to the fireplace and start my holiday reading
Tweet media one
1
0
11
@mslkmp
Max Langenkamp
7 months
The peaks in the cheese are about 7cm apart! Since the microwave is sealed, we know that these 7cm are from standing waves which are half the wavelength. 2.45E9 * 0.14 = 3.43E8 (c is actually 2.99E8 which makes this estimate within 15% - not bad!)
Tweet media one
1
0
12
@mslkmp
Max Langenkamp
1 year
Technical debt is a powerful metaphor for understanding our selves and our social systems: We are individually and collectively amassing technical debt all the time. 1/
1
1
11
@mslkmp
Max Langenkamp
2 years
@g_leech_ I put Phil Agre’s mail archive in a similar category — some of the best stuff really is not so easy to find
1
0
10
@mslkmp
Max Langenkamp
2 years
Lévi-Strauss (the foremost cultural anthropologist behind maybe Boas) was heavily influenced by linguistics and… Shannon’s information theory * Wiener’s cybernetics
Tweet media one
3
1
10
@mslkmp
Max Langenkamp
3 years
Me: “hey it’d be fun to hang out” Fellow grad student: “yeah! why don’t you look at some of my papers and see what we can collaborate on”
0
0
10
@mslkmp
Max Langenkamp
10 months
So pigeons probably do not sense north through flapping their wings (I was wrong) but instead through *seeing north* (h/t @km_xiang )
Tweet media one
@strandbergbio
Adam Strandberg
10 months
@mslkmp found in various places the stokes radius of calcium with hydration shell (a) and viscosity of cytoplasm (mu), with your 50 microtesla field value, gives essentially negligible velocity. Ca2+ radius is about 1 angstrom
Tweet media one
1
0
2
1
1
9
@mslkmp
Max Langenkamp
3 years
So many people excited about AI governance and crypto-govtech but have never been to their local city council meeting...
1
1
10
@mslkmp
Max Langenkamp
2 years
my master's degree is literally an NFT
Tweet media one
1
0
10
@mslkmp
Max Langenkamp
2 years
Taiwanese breakfasts are my favorite
Tweet media one
2
0
9
@mslkmp
Max Langenkamp
2 years
I was briefly part of a similar prestige manufacturing club at MIT — their only real function was to sell resumes as an appendage of the recruiting arm of financial trading companies.
Tweet media one
@ekzhang1
Eric Zhang
2 years
I wrote down a few words about school, my social environment, research, and a recent experience that made me think about this yesterday I honestly was kind of scared to share this publicly, but now I feel like I can, thanks @ronawang for the push
7
6
112
1
0
8
@mslkmp
Max Langenkamp
11 months
when I was in a highly uncertain/chaotic period in my life, Helen was one of the few people who made me feel heard and encouraged me to do the thing that felt good and alive. I remain very grateful to her
0
0
9
@mslkmp
Max Langenkamp
7 months
if you're doing a bunch of inferential statistics maybe your ontology is wrong
Tweet media one
2
1
10
@mslkmp
Max Langenkamp
7 months
am partway into writing my essay series on meaning: what it means to live in a meaningful world and how a technological orientation can, if we aren't careful, banish magic 1/
1
1
9
@mslkmp
Max Langenkamp
2 years
i have my first live gig this afternoon and i'm terribly excited. set list is autumn leaves, fly me to the moon, blue monk, blue bossa
0
0
9
@mslkmp
Max Langenkamp
2 years
Why doesn’t uploading your brain’s connectome to the cloud guarantee immortality? It turns out there is much more to your brain than neural connections.
Tweet media one
1
0
8
@mslkmp
Max Langenkamp
2 years
"technologies are artificial, but artificiality is natural to human beings"
Tweet media one
2
2
8
@mslkmp
Max Langenkamp
2 years
The full post is here: my guess is that my collection resources in one place, rather than my opinionated commentary, that is the most helpful (i.e. realizing there are dozens of wonderful different psychological tools you can choose from)
1
0
8
@mslkmp
Max Langenkamp
10 months
Just saw an old Taiwanese lady (I’d guess >70yo) doing pull-ups in a park. Have never seen anything comparable in the US
0
0
7
@mslkmp
Max Langenkamp
2 years
Applying for an ARPA-E grant right now and realizing how utterly utterly spoiled I’ve been by the new grant making institutions (eg @tylercowen ’s emergent ventures, @open_phil , @gitcoin grants…)
1
0
8
@mslkmp
Max Langenkamp
6 months
Thinking about a few dualities that can explain a lot about us and how we are in the world. Some scribbles:
Tweet media one
4
0
7
@mslkmp
Max Langenkamp
2 years
6. Finally, we make some (rough but falsifiable) predictions:
Tweet media one
1
0
8
@mslkmp
Max Langenkamp
2 years
Anyone in Tokyo wanna drink some nice genmaicha with me?
0
0
8
@mslkmp
Max Langenkamp
3 years
Broke: current bureaucratic research labs Woke: 70-80s era Bell Labs/Xerox Parc Bespoke: 0xPARC
@0xPARC
0xPARC
3 years
0xPARC is a new organization supporting decentralized application R&D. We aim to foster a budding community of researchers, developers, thinkers, and builders - a distributed "skunk works" for the Ethereum and decentralized tech ecosystems. (2/n)
2
4
64
0
0
8
@mslkmp
Max Langenkamp
3 years
when reading neurobio papers I occasionally come across sentences like this. they look like english but I have to convince myself that they mean something to somebody
Tweet media one
0
0
8
@mslkmp
Max Langenkamp
1 year
Tweet media one
@a_m_mastroianni
Adam Mastroianni
1 year
The fundamental problem with most attempts to reform science is that they want to turn the Big Ship. Instead, we should be launching lots of Little Ships.
5
24
119
0
1
8
@mslkmp
Max Langenkamp
2 years
Gratitude to @tuomasoi @SerenaLBooth @xuanalogue @phil_of_ai @cailean_osborne for sharp feedback, @hlntnr @timhwang @enfascination for early encouragement, @AdamSalisbury15 @nayafia for talking to me about their inspiring research, among others. Deep thank you all! 9/
3
0
8
@mslkmp
Max Langenkamp
10 months
Two questions: 1. Why do we have hair 2. Why are we the only large animals that can’t sense the earth’s magnetic field
3
1
6
@mslkmp
Max Langenkamp
1 year
Burnout is what happens when our bodyminds collapse under technical debt
1
0
7
@mslkmp
Max Langenkamp
1 year
I have to confess.. I actually am a @meaningness thirstbot. mainly on here for his tweets. not even joking
@Meaningness
David Chapman
1 year
👯‍♀️ Pretty sure Max isn’t a thirstbot. Otherwise, how about charging these girls a buck? I appreciate their enjoyment of my tweets but I think they should do their bit
Tweet media one
4
0
20
0
0
7
@mslkmp
Max Langenkamp
7 months
one of the habits that has given me most joy has been memorizing a new poem each month. my cousin and I started sending each other clips and it has made me love life more
Tweet media one
2
0
7
@mslkmp
Max Langenkamp
10 months
It’s remarkable how consistent the slower processing body-intelligence system is. I’ve been asking myself a question before sleep each night and the next day I have an often surprising answer before noon
1
0
7
@mslkmp
Max Langenkamp
2 years
(Feeling a bit vulnerable in sharing but warmed by the thought that even a single other person helped by this would make it worth it)
2
0
7
@mslkmp
Max Langenkamp
2 years
Meditation, at its best, has allowed me to fall in love with reality again and again
0
0
7
@mslkmp
Max Langenkamp
7 months
We know c = fλ (frequency times wavelength). Frequency of our Sunbeam microwave is 1.45 GHz (can verify using a spectrum analyzer), now we just need wavelength. What we're looking for is wavelength. Time to microwave the cheese (rotation paused)
Tweet media one
1
0
7
@mslkmp
Max Langenkamp
3 years
Even today, LM Sacasas and Nathan Schneider (both Christian afaict) are the critics of technology I find most thoughtful. They seem to spring from a similar tradition.
3
0
8
@mslkmp
Max Langenkamp
1 year
‘Dominated by clippy’ party at the media lab last Friday. The demon voice of clippy occurred to me in a dream and I produced the recording you can hear a snippet of in the second video
Tweet media one
Tweet media two
Tweet media three
0
0
7
@mslkmp
Max Langenkamp
1 year
To say that I am “in a bad mood” is in a way wrong. We think of moods as “inside us” and the world as “outside” but this ignores the fundamentally porous boundary between me and world. It’d be more accurate to say “the bad mood has me”.
2
0
7
@mslkmp
Max Langenkamp
3 years
Am a big fan so far of @yak_collective 's primer on how to do online governance. Exhibit 1: the @vgr signature 2x2 1/n
Tweet media one
1
2
6
@mslkmp
Max Langenkamp
2 years
Ok I’m gonna do this
0
0
6
@mslkmp
Max Langenkamp
2 years
To prevent next pandemic: Far-UVC light research, safeguards against DNA synthesis, more funding.
@bridgetw_au
Bridget Williams
2 years
Recently the World Bank announced a new pandemic prevention fund. While this is very welcome news, the $1.4 billion committed to the fund so far is a long way from its goal of $10b. This is just one example of underfunding of pandemic prevention. 1/12
1
11
56
2
0
6
@mslkmp
Max Langenkamp
2 years
I think we can trace a pretty much unbroken lineage from the transcendentalists (eg Emerson) to the hippie counter-culture to the 90s rave subculture to today
1
1
6
@mslkmp
Max Langenkamp
3 years
"AI safety is too wordcel for me" (overheard at MIT)
0
0
6
@mslkmp
Max Langenkamp
10 months
The MIT Media Lab's PhD requirements, especially its General Exam, are awesome. You choose three areas: Primary, Contextual, and Technical. 2/3 (Primary and Technical) usually involve science or engineering, while the Contextual area is explicitly non-technical 1/
1
0
6
@mslkmp
Max Langenkamp
1 year
got my Cambridge busking license
Tweet media one
0
0
6
@mslkmp
Max Langenkamp
1 year
Some privacy/security practices I wish had known earlier:
2
1
6
@mslkmp
Max Langenkamp
2 years
Partly for this reason, I think science in general will not be sped up very much by the most recent ML research. Link for my essay is here
2
0
6
@mslkmp
Max Langenkamp
1 year
The hottest source for US intelligence leaks
Tweet media one
0
0
6
@mslkmp
Max Langenkamp
3 years
Solarpunk exhibit 1 spotted at media lab
Tweet media one
1
0
6
@mslkmp
Max Langenkamp
11 months
US, China, UK, EU representatives just signed the Bletchley declaration acknowledging the potential for 'catastrophic' harm from AI models. There seems to be a slight 'AI safety race' dynamic happening between the USG and UKG
1
0
5
@mslkmp
Max Langenkamp
2 years
Full thank you list is going to come later but, briefly, my advisor @dhadfieldmenell and my collaborator Daniel Yue have been the scaffolding to my hectic ever branching thought-trees 2/3
1
0
6
@mslkmp
Max Langenkamp
1 year
How much of a cost do the Chinese pay for having a pictographic language?
2
0
6
@mslkmp
Max Langenkamp
2 years
Was fortunate enough to be in the initial screening of this. Luke's project is an incredible meditation on not just MIT but technological progress in the US since the start of WW2. Highly recommended
@lukeigel
Luke Igel
2 years
🎥 After 2 years of work, MIT: REGRESSIONS is now available to watch online. It's a feature-length documentary that spans 80 years of MIT's/America's history via AI-enhanced archival footage. Check it out for yourself, head to @mitregressions , read more for some fun details
5
26
188
0
0
6
@mslkmp
Max Langenkamp
2 years
@erikphoel You might get some mileage out of David Chapman’s latest online book ‘Better Without AI’. He clarifies for me the debates around sentience/agency and has my favorite measured take. My highlights:
@mslkmp
Max Langenkamp
2 years
Loving @meaningness ’s latest book on AI risk. I’ve been thinking about this since 2017 and the book clarified a lot of confusion I had, particularly around how we conceive of AI (though I have some objections too). Thread: 1/
1
3
13
0
0
6
@mslkmp
Max Langenkamp
2 years
5. Risk assessment: our trends point to concentration of capabilities within a few firms. This may be good if we want to reduce dangerous models release + believe in govt regulation, but bad if we are concerned with corporate accountability. 5/
1
0
5
@mslkmp
Max Langenkamp
2 years
Idea: for the most important classes you took in uni, pick an interesting question from the final and make a video explaining it to a younger version of yourself
2
0
5
@mslkmp
Max Langenkamp
2 years
Found in the archives
Tweet media one
1
0
4
@mslkmp
Max Langenkamp
2 years
Americans give much more of their personal money away than do other countries’ citizens . (total $580 bn in 2021). Many (eg Tooze) attribute this to high religiosity
Tweet media one
2
0
5
@mslkmp
Max Langenkamp
2 years
Sen. Johnson talks about how he was doing research on AI when he came across a Jordan Peterson interview on AI. [~14:30 mark] He then goes on to read excerpts from Peterson’s testimony for five minutes before arguing with the Chairman for another five minutes then leaving.
1
0
5
@mslkmp
Max Langenkamp
2 years
anyone have advice for what to do if you used your FTX grant funds to invest in crypto to pay off student loans? asking for a friend
2
0
5
@mslkmp
Max Langenkamp
3 years
When the most piercing critique of the CCP is from a Malay pop duo:
@SabinaKnight1
Sabina Knight 桑稟华
3 years
🧵~5.3 million views in 4 days. "Fragile"《玻璃心》mocks the oversensitivity of cyber-nationalists, internet trolls that guard #China against criticism. – by Malaysian #Chinese artist activist @nameweemusic 黃明志 (ft. Kimberley Chen 陳芳語)
Tweet media one
4
38
144
0
1
3
@mslkmp
Max Langenkamp
2 years
I guess I’m the of the new generation who has been exposed first to the reformed/more efficient grant systems and later to traditional grants
0
0
5
@mslkmp
Max Langenkamp
2 years
@hlntnr At the end of my thesis I made some predictions for 2023 re MLOSS both to put ‘skin in the game’ but also to encourage other people to do the same
@mslkmp
Max Langenkamp
2 years
6. Finally, we make some (rough but falsifiable) predictions:
Tweet media one
1
0
8
1
0
5
@mslkmp
Max Langenkamp
2 years
.For instance, some of the biggest misinformation problems in Brazil (election misinfo) and India (sectarian violence / COVID misinfo) happened on WhatsApp, which has no recommender algorithm. eg 9/
1
0
5