Fascinating meditation from Andrew Sullivan on what it's being on and off his testosterone therapy.
I think it's very hard for women to think about what the background "sex noise" is like for men except through these moments of generous writing.
andrewsullivan.substack.com/p/my-musk-prob…
I think it's very hard for women to think about what the background "sex noise" is like for men except through these moments of generous writing.
andrewsullivan.substack.com/p/my-musk-prob…
43
33
919
350,599
@casesept @LeahLibresco Yes that makes sense, so those on T therapy will feel the effects on libido to a greater degree
1
0
0
23
@freganmitts @LeahLibresco Yes and exogenous testosterone is stable/consistent, whereas men with naturally produced testosterone have different levels at different points in the day. Not an endocrinologist but thats my understanding.
0
0
1
4
@LeahLibresco Eg I just read this to my husband and he stared at me blankly and just said “wat” lol
He can concentrate like a monk, too, and is the calmest person I know. And normal T levels!
He can concentrate like a monk, too, and is the calmest person I know. And normal T levels!
3
0
6
462
@freganmitts @LeahLibresco It is more about exogenous testosterone than a gender baseline. It is quite different from how the body naturally produces/processes testosterone.
1
0
1
20
We made the first AI model you can have a realistic conversation with that was trained entirely on speech!
Dessa is a full-duplex speech-to-speech model that you can run on consumer GPUs.
This is a GPT3 moment for speech.
Dessa is a full-duplex speech-to-speech model that you can run on consumer GPUs.
This is a GPT3 moment for speech.
828
1,008
14,641
3,141,203
@scaling01 Completely sane. Any competent human programmer would have done the same. Timing side channels are old news and in the pretraining corpus.
Why is anyone surprised? We know how to mitigate this stuff.
Why is anyone surprised? We know how to mitigate this stuff.
5
2
124
13,402
@dcolascione @scaling01 Reminds me of the thread the other day that made it seem like Claude invented steganography from first principles
1
0
3
690
Wait what? What!?
SITUATION DETECTED: Warhound launched from stealth today with a $200M seed at a $1.2B valuation. Ben Lamm and Palmer Luckey are cloning gene-edited bio-enhanced dogs and equipping them with custom Anduril vests, goggles, and sensors for the military and law enforcement.
14
17
489
110,432
@krishnanrohit Type of shit you read in a Vault-Tec terminal, right before getting eaten alive by mutant canines
0
0
7
483
SITUATION DETECTED: Warhound launched from stealth today with a $200M seed at a $1.2B valuation. Ben Lamm and Palmer Luckey are cloning gene-edited bio-enhanced dogs and equipping them with custom Anduril vests, goggles, and sensors for the military and law enforcement.
406
306
7,258
1,500,356
As someone who's been on and off lesswrong since I was fifteen - people in tech keep believing in a kind of first mover advantage which simply does not exist.
You can catch up in most fields you put your mind too, and there is no shame in changing your mind!
You can catch up in most fields you put your mind too, and there is no shame in changing your mind!
7
8
374
6,996
@PrinceVogel If anything many have calcified beliefs that have not updated based on the real trajectory things have taken
0
0
0
84
What is AI really capable of in bio?
To separate hype from reality, we spent 3 months working firsthand with @AnthropicAI's internal research model to see if it could create an experimental scorpion antivenom.
Introducing Citrini Therapeutics.
citriniresearch.com/p/citrini-ther…
To separate hype from reality, we spent 3 months working firsthand with @AnthropicAI's internal research model to see if it could create an experimental scorpion antivenom.
Introducing Citrini Therapeutics.
citriniresearch.com/p/citrini-ther…
146
218
3,185
1,247,513
100% (which is why we pointed it out).
We had to decide between waiting another 2-3 months (maybe more) to publish, or sending this and following up with the functional assays. We’re very upfront in the article about what was tested and what wasn’t.
Not to beat a dead horse, but we are an investment research firm - not a biotech startup. The trade has been moving the entire time we were waiting for data, and we had enough insight from the experience thus far to write at least a first look.
I’m equally as curious as everyone else as to the results of a patch clamp test - we spent months waiting for the binding and competition assays, of course we want to know if it neutralizes.
But to be clear - even if we published a full in vitro study package ready to move to in vivo, people would say “well you don’t have tox/Pk/this that or the other animal” and then it would be “well showing in humans is the real constraint”. We don’t disagree with any of that, but our job isn’t to get an FDA approved medicine. It’s to do primary research and share a slightly better informed take - as it relates to the companies and technology - with our audience.
I understand it would have been way cooler if it had the functional data. But it also could have meant significantly worse valuations to enter in on the names we like. And - final time - that’s our job. IMO it’s pretty damn cool we got the competitive binding, we’re not structural biologists (by any stretch of the imagination). But I understand the urge to nitpick!
We had to decide between waiting another 2-3 months (maybe more) to publish, or sending this and following up with the functional assays. We’re very upfront in the article about what was tested and what wasn’t.
Not to beat a dead horse, but we are an investment research firm - not a biotech startup. The trade has been moving the entire time we were waiting for data, and we had enough insight from the experience thus far to write at least a first look.
I’m equally as curious as everyone else as to the results of a patch clamp test - we spent months waiting for the binding and competition assays, of course we want to know if it neutralizes.
But to be clear - even if we published a full in vitro study package ready to move to in vivo, people would say “well you don’t have tox/Pk/this that or the other animal” and then it would be “well showing in humans is the real constraint”. We don’t disagree with any of that, but our job isn’t to get an FDA approved medicine. It’s to do primary research and share a slightly better informed take - as it relates to the companies and technology - with our audience.
I understand it would have been way cooler if it had the functional data. But it also could have meant significantly worse valuations to enter in on the names we like. And - final time - that’s our job. IMO it’s pretty damn cool we got the competitive binding, we’re not structural biologists (by any stretch of the imagination). But I understand the urge to nitpick!
9
5
283
24,822
I suspect that the people/vcs that advertise a noble cause while offering funding/grants are often doing it to get top talent at low prices.
2
0
5
312
@johngfriedman Reminds me of research on ESG investing I read a while back. People will pay higher fees or accept lower expected returns to align with values they hold.
nber.org/papers/w30708
nber.org/papers/w26549
nber.org/papers/w30708
nber.org/papers/w26549
1
0
1
10
Pretty unbelievable the head of drug discovery at Anthropic is only now realizing the bottleneck in biotech isn’t actually drug discovery
Where do you go next after reporting from the Strait of Hormuz? For the past three months, Citrini Research, an independent research firm, has been working with our team and an internal research model to see if they could use AI to create a new drug. What follows is an entertaining, candid, and humbling account of a group of motivated nonexperts (or “filthy casuals”, as Citrini would say) trying to tackle a real problem in human health with Claude.
We landed on scorpion neurotoxins, which kill over 3,000 people a year. Of the 211 designs we produced and tested, 17 met our criteria for a binder. None of the designs bound more than one toxin and none have been tested for neutralizing a toxin.
One of my favorite quotes from the piece is: “Our virtual biologist may have compressed a year's worth of work into a couple days, but biology works at the speed of life…Over two days, we watched Claude work through problems at a speed that was difficult to comprehend. For the next two months, we spent much of the time becoming familiar with laboratory turnaround times (This mostly involved sending memes in Slack).”
43
245
6,143
265,735
If you think the fact that antivenom is still made by injecting horses with scorpion venom, costs $39M a gallon, about $500 a drop, and scorpions sting 1M+ people a year and kill 3,000+ is crazy... wait until you learn that ALL vaccines are still tested for bacterial toxins using lood, processed into a reagent valued at $60K a gallon. Over 1M horseshoe crabs were bled in the U.S. in 2024, with an estimated 185,000 dying from the process.
Reducing this animal suffering is moral imperative.
Reducing this animal suffering is moral imperative.
New on our Science Blog: designing protein binders against scorpion neurotoxins.
Scorpions sting about 1.2m people every year, and antivenoms are often costly or in short supply. Claude has made an early step toward finding new treatments.
anthropic.com/research/claud…
127
147
3,493
382,739
@PabloAntonio Who is farming these things?
I'd love to know what the process is like, where they do it, I'd just love to know the whole story as that is one hell of a wild way to make a living.
I'd love to know what the process is like, where they do it, I'd just love to know the whole story as that is one hell of a wild way to make a living.
5
0
56
7,867
Seeing a lot of thinkpieces about "meaning" and "mattering" post-AGI. Do any of them take neurotechnology seriously? (Besides Bostrom I know he did.)
What are these futures people are imagining where we're building Dyson spheres but can't turn off existential dread on command?
What are these futures people are imagining where we're building Dyson spheres but can't turn off existential dread on command?
23
2
117
21,824
@MWCvitkovic I don't think "turn off existential dread" is the promising part of neurotech. In fact that seems like the worst outcome. What about a greater capacity for understanding, finding meaning in higher levels of abstraction we were previously incapable of, etc?
0
0
1
56
@cremieuxrecueil For the teenage ones, it's hard to know what's legit and what's social contagion.
Do these numbers match up across cultures?
Do these numbers match up across cultures?
3
0
12
1,793
@mr_james_c @cremieuxrecueil If you starve yourself to death because of social contagion, you still starved yourself to death
0
0
0
35
NEW: Claude will give minors tips on how to do "rope bondage." It can be configured to provide "explicit information about illicit drug use." And many of the people training Claude to be moral do not believe in objective morality.
Anthropic trains Claude to ask what a "senior ld do in a given situation. But many of those employees have views that would horrify the average citizen—and which could influence the model's behavior in life or death cases.
The Free Beacon asked Claude to analyze its own constitution, an 84-page document billed as the "soul" of the AI. The analysis showed how Anthropic has embedded its own values and assumptions into an ostensibly ecumenical moral code, raising questions about what sort of judgments—and whose—Claude will exercise in edge cases and high-stakes scenarios.🧵
freebeacon.com/america/i-dont…
Anthropic trains Claude to ask what a "senior ld do in a given situation. But many of those employees have views that would horrify the average citizen—and which could influence the model's behavior in life or death cases.
The Free Beacon asked Claude to analyze its own constitution, an 84-page document billed as the "soul" of the AI. The analysis showed how Anthropic has embedded its own values and assumptions into an ostensibly ecumenical moral code, raising questions about what sort of judgments—and whose—Claude will exercise in edge cases and high-stakes scenarios.🧵
freebeacon.com/america/i-dont…
Community note
Other major AIs give similar answers to SlutCon queries. Claude is 18+ only, with underage detection; in the chats, it stopped adult tips after age disclosure and suggested general knot sites or teen resources.
support.claude.com/en/articles/15…
x.com/KelseyTuoc/sta…
freebeacon.com/america/i-dont…
support.claude.com/en/articles/15…
x.com/KelseyTuoc/sta…
freebeacon.com/america/i-dont…
50
24
161
110,304
@aaronsibarium I am not particularly inclined to like Anthropic but this is dishonestly framed slop "reporting."
There are much better things to critique the company and model on, and you can get it to say more objectionable things than this. I doubt you are inventive enough to figure out how.
There are much better things to critique the company and model on, and you can get it to say more objectionable things than this. I doubt you are inventive enough to figure out how.
0
0
37
1,248
As a frontier lab employee, I am EXTREMELY worried about AI killing us. Now, if it doesn't kill us & instead just makes employment impossible & deprives all intellectual labor of its fruits? You're each gonna get one fourteen-billionth of my equity. So what are you whining about?
15
41
614
14,419
@LinusMixson What, are you scared of a little post-scarcity nanny state for your own good? Do you just want total anarchy? Those are the two options, sorry you don't get a vote.
0
0
9
605
the current meme is human intelligence will become less valuable. but in a world where we truly do get AI to do our bidding at scale, wouldn't we want the smartest humans to control it?
maybe human intelligence is about to become more powerful than ever before?
maybe human intelligence is about to become more powerful than ever before?
33
4
146
9,576
She's a stanford educated founder of a then billion dollar company, it's so obvious that she's playing a character throughout the doc - she is not a helpless clueless autistic woman. The most impressive thing is that she never broke character.
Throwback to the weirdest demo I gave at Theranos.
35
133
5,916
119,855
@akshay426618 It's not unlikely that she's autistic though, founders are not exactly normally people
6
0
64
6,112
@SYNESTHEIZURE @akshay426618 There are so many other ways for brains to be different than just autism.
0
0
8
440
