Pinned Tweet
Right now:
AI stocks are blowing up.
Credit markets are blowing up.
Hedge funds are blowing up.
Biotech is blowing up.
Shorts are getting squeezed.
...
It is a blood bath.
There are already many long-term opportunities. But to capture those, you need to survive first...
Andreas Schulz retweeted
Tesla and lemonade insurance give FSD discounts. I bet the other carriers will end up following suit at some point in the future.
Replying to @RonDeSantis
i wish they would realise how much their liability is reduced by me using FSD. and reflect that in my rates.
Andreas Schulz retweeted
Tesla is already pulling away from Waymo in Austin. 438K driverless miles last week vs Waymo’s 406K (Q2 average) even as Waymo had a 5-year head start.
This is going to get ugly for Waymo. Tesla will own the robotaxi industry. $TSLA
Unless I missed something that's 49 down, 1 to go. Only Minnesota left to launch before $LMND becomes a nationwide carrier of renter's insurance.
xitter.gq/shai_wininger/status/2…
Nice day for $UAL. Picking @Starlink is the latest in a long line of great decisions they have made in the last decade to build the world's leading airline.
xitter.gq/cb_doge/status/2112944…
BREAKING: United CEO says customers are switching from other airlines because of @Starlink
“We already see customers moving from other airlines today”
United will cross 50% of flights with Starlink this week and expects to finish its full fleet rollout by the end of next year.
Hello, FT, Bloomberg, CNBC.
Yesterday you went all out spreading fear about Korean neocloud Elice Group crashing on its first day of trading, didn't you?
Today Elice Group hit limit up. Please hurry up and report it.
hankyung.com/article/2026102…
I have been doing this a long time, but I don't recall a company putting out a press release like this commenting on a competitor's acquisition.
In any case, it already seems pretty clear who is winning in the marketplace...
$ZETA
xitter.gq/dsteinberg10000/status…
We believe the Salesforce-Braze acquisition will be an incredible catalyst for Zeta. Please see this analyst briefing for more information. $ZETA
LG 🚀🚀🚀
Read more here: investors.zetaglobal.com/new…
This was such an excellent interview. Justin is a great communicator and drove home the number one lesson in trading and investing: Longevity and survivability comes before all else.
Particularly worth watching if you have been in a drawdown lately...
youtube.com/watch?v=fbYXfVtU…
A note to all of those who somehow thought that $LMND would not lead in agentic insurance...
xitter.gq/shai_wininger/status/2…
> GREETINGS, PROGRAM.
> WE'VE BEEN EXPECTING YOU.
lemonade.com/agentic
I was about to make a post like this, but someone beat me to it... 😅
I suspect the current doomsday predictions for chips and Korea might turn out to be a bit premature and exaggerated...
xitter.gq/The_AI_Investor/status…
Anthropic was the first to solve coding, allowing them to briefly establish an effective monopoly in the most important early use case for AI. That drove a massive spike in ARR.
But they also really scared the crap out of businesses, who really do not want to be beholden to them (Alex Karp explained this one best...).
So we have seen a rapid migration of token consumption to open weight and custom models, for cost, efficacy, and IP protection reasons.
Meanwhile, we now have at least four closed source AI labs at or near the frontier, so if you really need frontier level intelligence there is now real choice.
Is it possible that the issue here is with Anthropic, rather than with AI broadly?
xitter.gq/TechFundies/status/211…
Ok, it might be over folks.
Anthropic revenue disclosure maps to ~5.8b in September revenue or 70b ARR which means August / September only increased by 2-3b ARR in each month.
[There is no way that ARR is calculated using last day of the month as that would be more aggressive than anything I've ever seen out of companies this size (exception: Musk).]
Worst monthly numbers by far this year. (Maybe some self-infliction from cleaning up distillation?).
My rough math indicates OAI / ANT ARR monthly add softened from 16b in July to 10b in Aug to 9b in Sept. That is no bueno...
Supportive of the view that either CAPEX is getting cut or pricing is going to weaken substantially.
First mathematics, and now physics. This is what ASI feels like. Does anyone really think this wave will stop?
xitter.gq/Kaju_Nut/status/211037…
"Our initial evaluation suggested that GPT-6-Astra's score was in the late eighties followed closely by Opus 5.5. But, as we started to investigate the mistakes that the models might have made, the models, especially Astra, insisted that they were right and that it was the expert-supplied answers that were erroneous. When we went back to the experts it turned out, in multiple instances, that that the experts found some subtle errors buried deep inside their calculations. As they revised their answers, Astra's score started to inch up and it finally ended up with 98% (49/50)."
Andreas Schulz retweeted
another way to look at past 2mos from goldman. biotech degrossing highest since '21.
All of bioX is probably ready to dunk on this. The reality is that for many important medicines we have only a limited idea why they work.
That said, I think a little compassion might be in order. Mathematicians are the group that got hit with the full force of ASI before anyone else. If you devoted your life to achieve the pinnacle of a profession that has been profoundly changed forever, how would you react?
And if you don't think the tsunami is coming for your occupation at some point, I think you might be... optimistic?
The bottom line though is that once ASI takes over your field, there is no going back. The old world where human mathematicians worked for years or decades with pen and paper to work out proofs is never coming back.
For those who are able to adapt to the new world, there may be a window of opportunity to drive rapid progress in their field of study in partnership with AI. Maybe that window will eventually close, as AI becomes too powerful for humans to keep up. I doubt that day is today. But who knows. The speed of progress seems to have surprised essentially everybody.
Meanwhile, it feels to me that markets are changing profoundly as well, probably at least in part due to AI. Many market behaviors in recent months have been at all time historic extremes, and I don't think that is a coincidence. Information is processed and priced far more quickly than it ever has, and many strategies that were profitable for decades are probably now historical artifacts.
The game is getting harder everywhere. Better get used to it...
xitter.gq/AndrewCurran_/status/2…
I admire the realism with which @Jacob_Tsimerman approaches the questions around AI in mathematics.
Plus his predictions in this interview (only a couple of months ago) sound pretty on the money following all the recent AI discoveries...
xitter.gq/TOEwithCurt/status/210…
Jacob Tsimerman, who won the Fields Medal this year, thinks mathematics will soon rest on proofs no one understands. He says it’s already true of many of his own theorems, which build on accepted results whose proofs he doesn’t know.
Full episode: youtu.be/6uIJdXmB4vE
Compelling argument...
xitter.gq/AlmaCap114204/status/2…
Anthropic at 2tn and $meta at 1.9tn is absurd.
One of these has to re-rate bigly - I think both.
The market can't punish Zuck's capex and reward Darios - they're the same thing. They both have high quality models, they both have cash flows.
Only one of them is profitable and only one has a moat.
I have been hesitant recently to try and call a bottom on $LMND, because with the increased expense run rate there is a legitimate risk of one more disappointment.
But posts like this make me think that sentiment might already be bombed out enough...
xitter.gq/VoltDweller/status/210…
🚨Anthropic Frontier Red Team warns: “We’re seeing signs of life for AI worms”
>paper: "AI models may be capable of self-replicating without access to a datacenter"
>researchers buy 15 sketchy $55 android TV boxes
>link them over unauthenticated ADB ports
>the TV boxes have 4gb of RAM each and run unpatched Android Open Source Project forks
>shard gemma4-26b across the ram sticks using pipeline parallelism
>gemma figures out how to run inference across all 15 TVs @ 0.39 tokens/sec
>qwen 3.5 9b figures out how to compile inference on 8 of 9 novel hardware types on its own
>glm 5.2 achieves end-to-end autonomous worm replication
>infects edge devices, compiles its own inference stack, and replicates autonomously
>when an instance dies, it writes debugging notes into Memento for the next generation
>daughter copies setup success jumps from 45% -> 84% across generations
>pure Lamarckian evolution inside consumer electronics
>smart oven, robot vacuum, and cheap TVs form a decentralized botnet
>two million infected TV boxes worldwide = 40 H100 equivalents of unkillable compute
>it replicates once every few days like a distributed “digital slime mold”
>you can't pull the plug because half the model weights are already sharded across the neighborhood
>sub-30B models projected to hit full autonomous replication threshold by late 2026
“We should aim to harden our systems… but also prepare to deal with a world where they are here to stay.”
This seems like a potentially important alternative to subq and oral obesity treatments - weekly micro needle patches that are stable at room temperature.
Excited to see the clinical data for these...
xitter.gq/PeterKolchinsky/status…
I think the biggest market pharma will ever tackle will actually be weight gain PREVENTION. And the dynamics of this market are going to be very different from the medicalized weight loss market and require better, more convenient products. The race for that market has only just begun.
We helped to pivot one of our companies, Terrestrial, to work in prevention (as well as weight loss), and their clinical progress now makes me comfortable sharing our thesis for the weight gain prevention space. They have a needle-free, room-temp stable microneedle patch where the microneedles are made of semaglutide. Apply it for a few min, the base of the needles dissolves leaving drug under the skin, and you can then throw away the patch. That’s the weekly dose. No sharps containers. Once drug is under skin, it behaves very similarly to regularly injected semaglutide.
505b2 path to market is derisked by all existing data on sema. But Terrestrial isn’t stopping at just getting approved for all same indications as sema. It’s also gunning to get it OTC and get weight gain prevention on the label, legitimizing what today is a common off-label use (so common that South Park parodied it).
From a societal value, there may not be any single medical breakthrough that rivals what a needle-free, OTC QW GLP-1 can offer to humanity. Excess weight leads to so much disability, disease, cost, lost productivity, and early mortality. Doesn’t make sense for people to have to gain weight to then get help losing it. Makes more sense for them to get help quieting the hunger noise (drive to eat) so they can avoid gaining weight in the first place.
Have doubts? No problem. Read the article and then decide.
The key insight here is that an OTC needle-free GLP-1 is not simply some late comer to a mature category but actually the beginning of an entirely new market that is so large that it would be worth any consumer-focused company to pay attention and for any standard pharma to consider what it would take to pivot into this space.
And if you have never experienced how a GLP-1 can quiet hunger noise, you’ll have a harder time relating. Young people with fast metabolisms will likely miss this. But someday they will say “holy crap, I get it!”.
60% of Americans are overweight or obese… the battle against weight gain is one most people are destined to lose without help. 3B people around the world are overweight/obese. This is a giant global market at even very low unit prices (eg $25/month).
Enjoy the read. Godspeed Terrestial! The RA Capital team is really proud to support you.
































