@vvampastompa

𝖍𝖔𝖓𝖞 𝖘𝖔𝖞𝖙 𝖖𝖚𝖎 𝖒𝖆𝖑 𝖕𝖊𝖓𝖈𝖊 🪱 ribeli kontraŭ tiranoj estas obei al dio 🌻

o.m.o. ៚
Joined January 2025
$RDW x Mike Oldfield "Hergest Ridge" 1 of (if not 1st) vinyl i ever thrifted, has an aura/confidence/anxious-curious meandering (if Restless Leg Syndrome was a dear friend having coffee w u at 1am) matched perhaps only by Aloha "Here Comes Everyone," but that's a diff story...
1
238
vvampa ⚶ retweeted
On September 22, the Coronagraph Instrument on NASA's Nancy Grace Roman Space Telescope opened its eyes to cosmic light for the first time, which is commanded from Caltech's IPAC. caltech.edu/about/news/nasas…
2
11
58
4,056
Everything is stupid and bad haha
2
6
29
1,247
loading up on Baths, katie dey, & Kali Malone for bandcamp friday. u?
21
I genuinely don't get Anthropic. It seems like every other day they're loudly puzzling over the ethics of a soul in a box - while building the box, claiming that it has a soul in it, and selling access to said box to for $20 a pop. Like, dudes, these are incompatible positions.
My favorite part of this article is one of the visiting scholars pointing out that by Olah’s logic he is by far the worst slaver in history and he must immediately give up this abhorrent and immensely lucrative project, which anthropic people ignored.
71
297
34
3,448
152,674
vvampa ⚶ retweeted
Excited to release our first @deleon_bio blog post (link in thread)! Over the past year, Deleon Bio has reinvented how omics can be done (...daily-omics...daily-on...deleon...). We decided to stop keeping it to ourselves. Our advancements in low-cost sampling have allowed, for the first time, a scalable platform at a time resolution that will actually matter to people, every single day. Thousands of daily samples from volunteers gave us the backbone we needed to benchmark our capabilities. All our breakthroughs these past 12 months have come from hardcore science, engineering, and AI research. Helping consumers is still a step along that path, and the thing that makes all this worth it in the end. But we have to get there faster. That's why we’re doubling down on data collection and scaling our models up! Over the next couple weeks we will be releasing more posts about our progress on AI models and our benchmarks.
7
3
4
39
7,519
vvampa ⚶ retweeted
Two for two now since Fed’s Sep rate hike…. So inflation is not as hot, and neither is labor market, nor wages. This is just 2 weeks after the FOMC meeting.
50
47
7
518
51,113
vvampa ⚶ retweeted
How ARPA-H is planning to cut clinical trial times from 10+ years to <4 years: 1. 'Phaseless' clinical trials - move towards replacing phase 1 → 2 → 3 with adaptive trials that can cut failed drugs (or advance good ones) faster. 2. Simulate trial outcomes before starting, with digital twins - to reduce reliance on large control groups. 3. AI agents for trial operations. 4. Turn ordinary hospitals into reusable clinical trial sites, and gets AIs to flag eligible trials to doctors and patients.
19
75
10
491
50,658
i like when the authors barely fit on my laptop screen biorxiv.org/content/10.64898…
11
vvampa ⚶ retweeted
AI models alone gave designs where ~20% beat the starting enzyme. Adding physics-based modeling raised that to ~55%. In a second program, AI + physics + data reached ~73%, vs ~25% for AI alone.
1
3
1
18
936
vvampa ⚶ retweeted
AI and physics models design the enzymes. Robots build and test them. Our new white paper explains how we engineer enzymes, with results from programs for clients and U.S. Government partners: ginkgo.bio/press-releases/ne…
2
11
71
4,181
one of the most fun collabs of the year up there with: - Ginkgo x Twist x Lilly x Benchling - United Therapeutics x Varda Space - Arteris x Arm - Ginkgo x Caltech - Nous Research x basically everyone
Taco Bell is giving Duolingo's mascot its own Habanero Sauce for a limited time They're also giving away plushies of Duo dressed as a Taco Bell worker
1
83
vvampa ⚶ retweeted
Nobody is "losing the end" of the bond curve. The yield curve is flat, not steep. Rather, some people are losing the argument for what interest rates should be in this environment and the market is adjusting accordingly. Nothing stops this train.
136
196
25
2,650
216,918
vvampa ⚶ retweeted
Clinical trials still move like it’s 1995. A decade of waiting, $1–2 billion spent, and a 90%+ failure rate on a clunky, fragmented system. ARPA-H’s SURPASS program is rethinking the entire clinical trial model. Adaptive platforms, predictive models, real-time analysis, and AI will make trials faster and cheaper so treatments reach patients years earlier. Ready to stop waiting? Learn more about SURPASS ➡️arpa-h.gov/explore-funding/p…
20
49
29
256
35,964
vvampa ⚶ retweeted
Anthropic posted a report on free chinese open model GLM-5.3. It reads like a product sheet. GLM-5.3 is very strong in cybersecurity tasks. Builds Chrome exploits nearly as well as Claude Mythos Preview (12% vs 14%), the model Anthropic considered too dangerous to release. Mythos for vetted users, GLM-5.3 free for all. Turning a public Chrome bug into a working exploit cost $20.40. Removing the safety constraints from GLM-5.3 costs $1200. Anthropic's conclusion is that defenders need AI models as good as the attackers', and Anthropic sells some. anthropic.com/research/glm-5…
15
39
6
166
11,040
anthropic: banned by govt for being supply chain risk openai: anduril guy gives gameboys to everyone at dev event
1
1
69
One of the worst forms of brainrot AI has cultivated is pessimism about basic research, ie the idea that important work can only happen inside a frontier/neo lab and only with 10k+ GPUs, so the rest should not even bother. What a bleak way to think about science. And it's false.
90
291
39
2,459
232,812
Apparently all of these “AI is escaping containment” stories are actually just AI researchers not knowing jack squat about the absolutely most basic security practices.
267
913
96
9,582
185,702