Synthetic.com (@syntheticdotcom) header
Synthetic.com (@syntheticdotcom) profile picture

https://nitter.gq/t.co/XTgY1wsRFg a source of insights for all things biotech. Publisher & editor @slchase

Joined June 2020
Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)Photo by Synthetic.com (@syntheticdotcom)
AI's need for simulation and synthetic data grows daily. This applies to training to improve AI inference and robotics' training in simulation. Own the best piece of internet real estate synthetic.com
4
6
92
Synthetic.com retweeted
frontier models, with lots of test time compute, are becoming the engine of scientific discovery
416
404
151
7,510
575,741
Synthetic.com retweeted
Replying to @nicbstme @levelsio
I designed this small space with AI. Drawings, materials, supplies (link to specific items to order) and hi-rez color mockup. Instructions to carpenter in Spanish
1
3
7
169
Synthetic.com retweeted
If you’ve spent your career in DC accumulating power and influence there is virtually no chance you are not a psychopath – at Manhattan, NY
3
7
Synthetic.com retweeted
In 2024 a giant liability will mingle with a giant debt
1
5
1
6
Synthetic.com retweeted
Cuba’s Tourism Guest Book Is Becoming a Family Register cubajournal.com/travel/cuba-…
1
3
46
Synthetic.com retweeted
Yes, AI is rapidly advancing mathematics (and theoretical physics problems that are essentially mathematical). Like Ramanujan’s notebooks, human mathematicians will spend months (years?) poring over the stream of papers that AI effortlessly generates, hoping to verify and understand them, probably until math itself (via Lean) becomes a branch of computer science like almost everything else. However, after math and physics, there is still a next scientific frontier: chemistry, and especially biology and medicine. I mean, it was already the frontier, but it’s the next one too. AI can crush deterministic math problems on the computer, but biomedicine is a highly empirical, noisy, multivariate, and yet often predictable domain. Thus, AI radically increases our ambitions in biomedicine. All the omics data, all the messy literature, it can all be synthesized and analyzed in a flash. This is what the labs mean when they talk about curing all diseases. The tooling is available to take cracks at biomedical problems previously thought impossible. Midjourney is doing this, Chan Zuckerberg is doing this, Anthropic is doing this, but there will be so much more.
188
157
37
2,071
181,513
Synthetic.com retweeted
When does AI take off? When open-weight models running on bankruptcy-priced compute. Creditors-turned-owners have every incentive to run flat out at anything above power and opex rather than wait for a recovery.
3
5
87
Synthetic.com retweeted
Retatrutide has an advantage over other GLP-1-based drugs, and I think it comes down to glucagon, one of the more overlooked hormones in metabolism. Retatrutide adds glucagon on top of the GLP-1 and GIP signals. In the newly-released phase 3 TRIUMPH-1 trial, just published in NEJM (NCT05929066), adults with obesity took it for 80 weeks. At the top dose, people lost 25% of their body weight on average. More than a third lost at least 30%. Fasting insulin fell by more than half, something I always focus on, and over 90% of people with prediabetes ended up with normal blood sugar. Insulin tells the body to store fuel, and glucagon tells the body to burn it. In a separate study, researchers raised glucagon to normal physiological levels and found that the liver's mitochondria burned fuel 50 to 75% faster (PMID 39197461). In another study, a glucagon infusion raised energy expenditure by about 15% (PMID 26434748). Ideally, if you take it, protect your fat-free mass with resistance exercise and abundant protein and natural fats. I also feel strongly that any of these drugs needs to be viewed as a temporary crutch--use it to develop new habits (i.e., control carbs) and then taper off.
44
26
2
385
22,056
Synthetic.com retweeted
White House TLT buying today, before the great yield curve control announcement on Monday morning?
2
3
76
Synthetic.com retweeted
Replying to @rcwhalen
Rates of Hormuz
4
6
117
Synthetic.com retweeted
A couple small cap biotech names I believe are undervalued and under followed with potentially re-rating catalysts in the next 3-6 months. Both companies are funded well past these potential inflection points. NFA $KZIA - $180 mil market cap, $100 mil EV Lead Asset: Paxalisib - PI3k/mTOR dual inhibitor. Brain penetrant. Well tolerated. Very strong early signal in TNBC & expanding into HR+ and CRC. Robust depletion of CTCs, CTC clusters, and Terminally exhausted T Cells suggest that Paxalisib reinvigorates the immune system. 83% ORR in initial TNBC cohort. Very hot space with $CELC and $FTH trading at $4.5 bil and $1 bil respectively. KZIA lead asset is only ~1 quarter behind FTH lead asset in HR+ breast but trades at $760 mil less enterprise value. I believe this gap should start to shrink as KZIA publishes more data across their breast cancer clinical trials. Catalysts: Dec - San Antonio Breast, update on new and existing TNBC patients Mid-year 2027 - Initial HR+ dataset $EPRX - $545 Market Cap, $412 Mil EV Lead Asset: EP-104GI - Extended Release Fluticasone targeting once per year treatment of Eosinophilic Esophagitis (EoE). Open label data to date has been quite strong showing similar efficacy to Dupixent. I believe a once per year dosing schedule with full year suppression of EoE symptoms could present a potential blockbuster opportunity for $EPRX and would be used before Dupixent in most cases. With the strength of Dupixent launch in past 18 months, EoE is hotter than ever and the market should reward a positive phase 2 here. Catalysts: September - 12 month open label symptom data December - 6, 9, and 12 month data from randomized EoE trial
2
7
2
39
7,721
Synthetic.com retweeted
Why is it such a leap to understand that all grand threats of regulation and/or investigation are nothing more than lead gen for campaign contributions (bribes)? And this is political party agnostic
4
6
152
Synthetic.com retweeted
If open models were actually as risky as the frontier labs claim, they wouldn't need a PR campaign to make the case. So what's the real grift budget behind this fear campaign? Cash and compute both.
1
1
87
Synthetic.com retweeted
Quick numbers that still get me: $10,000 invested in Monster Beverage stock on January 1, 2000 would be worth roughly $21 million today. Back then it traded around 4 cents a share, split-adjusted, still going by Hansen Natural. In 2013, I wrote a piece calling Monster's business model unredeemable, comparing its products to derivatives "designed to fail": simonschase.co/sodastream-vs… — At the time its stock had already gained 20,318% since 2003, more than triple Apple's gain over the same stretch. I was right about the ethics. I was wrong about the trade. The stock kept compounding another decade past my skepticism.
4
5
128
Synthetic.com retweeted
Replying to @TheStalwart
$1 trillion in administrative overhead. Narry a word about using AI to fix this
If you can understand that obesity is a form of starvation, then you will grasp the current state of the US economy
2
7
1,218