Massive thank you @generalcatalyst, @graylark and the Green Berets, SEALs & Rangers for an incredible morning at the range.
๐ค Made with AI
sam: local inference is a terrible allocation of scarce resources
โa lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of thatโ
RIP Patrick Stewart. I will never forget watching @StarTrek TNG as a kid. Role model and deep inspiration for many techno-optimists. Make it so.
2
167
10/10 event organizer
Looking at hosting more events through out the country due to the success of our hardtech meetup in SF. We had over 80+ founders/VCs come; many YC founders as well as @a16z and @sequoia presence.
Currently have opportunities to host in DC, FL, SF and NYC coming up with sponsorships now.
Anyone interested in hosting/sponsoring a hardtech/defense meetup in their city?
Purpose? Meet some cool people in hardtech/defense and host a fun party/event.
I have been hosting events at mansions, yachts and (most recently) a robot fighting arena.
What I provide: I'll bring everyone together and ensure everything goes smoothly on time. Everything else can be figured out if I have a venue/interested sponsors in a popular city.
Comment on this post/shoot me a DM and we'll figure out dates.
love this. Generative is so compelling clear upside, but we shouldn't sleep on predictive AI.
JUST IN: Tractian raised $130M at a $1.4B valuation to build sensors that warn factories before their machines break down.
Its sensors attach to industrial equipment, tracking vibration, ultrasound, and other signals to detect problems early.
-AI identifies faults and recommends fixes
-Trained on a massive industrial failure dataset
-Used in 2,000+ facilities worldwide
-Customers include Coca-Cola, Whirlpool, and Eli Lilly
-Also monitors data centers and critical infrastructure
Unplanned downtime costs the worldโs 500 largest industrial companies an estimated $1.4 trillion annually.
2
146
๐
robots need control panels too
engineered and built this one for @aseonlabs robotic AV servicing pod
1
2
133
Robot fights + incredible builders ๐บ๐ธ ๐ฆพ
Huge thanks to @jjkaplowitz, @natabr75, @edgarpavlovsky, and @nickco for organizing, and @REK for hosting.
SF hard tech goes HARD
Shoutout to @jpmorgan for investing in grid resilience
Hardtech founders they're worth a look
jpmorgan.com/insights/securiโฆ
102
๐บ๐ธ
LIVE: @SecWar speaks to employees at Riverside Machine and Engineering during an Arsenal of Freedom tour stop in Eau Claire, Wis. xitter.gq/i/broadcasts/1qGoNYDlXโฆ
48
hell yes
We started Arda in 2025 to speed up American manufacturing. We have spent the last year building with Anduril as our design partner. Today we are launching the company publicly on CNBC and announcing our Series A: $160M at a $1.5B valuation.
Our platform helps manufacturers design and operate manual assembly lines, putting designs into production faster now. Next we'll build automated lines that can be reconfigured as quickly as human workers, letting America make products that now require armies of workers in Asia.
3
127
What happens to AI-dependent critical services and institutions when the cloud can't be reached?
As AI becomes essential to more services, infrastructure must be designed to keep critical functions running even when connectivity fails.
More from @jase on what it takes to build dependable infrastructure. ๐
ready.net/blog/what-happens-โฆ
2
90
Big movement in hyperscale data center debt this week. It seems we're entering the first serious stress test, and deterioration has accelerated sharply:
1. CoreWeave is flashing a serious credit warning.
Its bonds have sold off for two consecutive sessions, pushing yields above 15%. Its shares fell approximately 15.5% yesterday.
CoreWeave matters because it's a highly leveraged AI infrastructure operator whose business depends on large capital expenditures, long-term customer demand, and continued access to refinancing. Credit investors are now questioning that combination much more aggressively.
2. Major data-center fundraising plans are being postponed.
Reuters reported this morning that proposed listings involving Blackstone-backed AirTrunk, Nvidia-backed Firmus, and SoftBank-backed SB Energy have been pulled or paused. Firmus reportedly struggled to justify ambitious projections alongside approximately $30 billion of debt.
Investors are no longer accepting projected AI demand as sufficient justification for valuations and leverage.
3. Interest rates are pressuring the entire development model.
The 10-year Treasury reached roughly 5.22% yesterday, compared with 4.75% in August. It briefly exceeded 5.30% following Wednesday's inflation report.
Even before lenders increase their credit-risk premiums, the underlying cost of capital has moved substantially against developers.
4. The newest warning: default insurance is exploding
What do the credit insurers think? This morning's Reuters credit-market report is particularly concerning.
CoreWeave going north of 1% on CDS 5 year spread means insurers are more convinced than not that it will default within the next few years.
reuters.com/commentary/reuteโฆ
1
151






















