Mobitz44
124 posts
BREAKING: Early reports disclose the U.S. has spent $2 trillion worth of military ammuniton and equipment supporting allies in the first 45 days of the Iran war.
English
@Soulbagalma @assface_private Trade of the year for me (so far…)
Bought 300c’s in the 295-310 range, sold at 390.
English
@assface_private I really liked your avgo. I think it went down to 5 and closed 79+. But I closed too early on the initial run up. We were too early but if you averaged in, it was a generational trade. Not sure if anyone held. This is for April but if you got the May you would have made 💰💵
English
@assface_private It would be borderline negligent to not have at least some short exposure over the weekend right now.
English
@assface_private @byehiselllo People love to blame everyone but themselves for their inability to hit the sell button at the appropriate time.
English
@assface_private I think the 5% of port 1 dte 7100c was the correct call at the time
English
@PapaVentura7 @assface_private Other way to look at it: the way the market dipped on rumours of negative news was drastic. Did it in overnight as well. It will happen again.
I agree that timing of entry short here is really hard though.
English
@assface_private The way that dip was bought though…market might never go down again
English
SpaceX Cursor deal is Musk chasing losses. Surprisingly shortsighted move from someone that favours a first principles approach.
I feel this stems from either frustration of shortcomings of xAi/current models or a lack of understanding of the clear asymptotic nature of current approaches.
Cursor acquisition solves nothing. 🤷
English
@IntCyberDigest Guy cracked under zero pressure. 50 hour week 😂 this guy French or something?
English
Oh my, if you're having a bad day you should look at this person's day. 💀
English
The costs of striking Iran are real. But so is the nuclear threat.
Iran entered 2026 with enough uranium for 10 nuclear bombs. Before the June strikes, it was days away from enriching enough for 1 bomb—a level far beyond plausible civilian needs.
Operation Epic Fury is working.
We are systematically dismantling Iran’s war machine: missiles, drones, air defenses, navy, nuclear sites, defense industry, proxy networks, central command. In under 3 weeks, the supreme leader is dead, his successor wounded, and Iranian ballistic missile and drone launches are down 90%+. Iran is losing capacity faster than it can create chaos.
"War is never clean. But the strategy—the actual strategy, measured in degraded capabilities rather than cable news cycles—is working."
Excellent article.
aljazeera.com/opinions/2026/…
English
@assface_burner Eh, usa still can't pass. This is nothing burger
English
@cryptopunk7213 This is complete bs. H100 costs are 1/3rd of what they were 2 years ago. Holy shit people just take anything said on a podcast at face value. Not a brain cell in sight.
English
this completely fucking breaks the AI Bubble narrative.
a 3 year-old gpu is MORE valuable today because it serves higher-quality ai tokens FOR CHEAPER.
translation: gpt 5.4 runs BETTER on an OLD GPU than gpt-fucking-FOUR
read that again.
a newer, better model runs more cheaply on a shittier GPU
if this trend continues you can run claude 6 or gpt 6 on a fucking napkin
the trend is INVERSE.
Dwarkesh Patel@dwarkesh_sp
.@dylan522p gives a deep dive on the 3 big bottlenecks to scaling AI compute: logic, memory, and power. And walks through the economics of labs, hyperscalers, foundries, and fab equipment manufacturers. Learned a ton about every single level of the stack. 0:00:00 – Why an H100 is worth more today than 3 years ago 0:24:52 – Nvidia secured TSMC allocation early; Google is getting squeezed 0:34:34 – ASML will be the #1 constraint for AI compute scaling by 2030 0:56:06 – Can’t we just use TSMC’s older fabs? 1:05:56 – When will China outscale the West in semis? 1:16:20 – The enormous incoming memory crunch 1:42:53 – Scaling power in the US will not be a problem 1:55:03 – Space GPUs aren't happening this decade 2:14:26 – Why aren’t more hedge funds making the AGI trade? 2:18:49 – Will TSMC kick Apple out from N2? 2:24:35 – Robots and Taiwan risk Look up Dwarkesh Podcast on YouTube, Apple Podcasts, or Spotify. Enjoy!
English
@sarahdingwang Just because it’s said fast and with confidence doesn’t make it correct. H100 cost/hr is significantly cheaper today than 2 years ago.
English
Narrative violation from Dylan on Dwarkesh: H100s are worth *more* today than they were 3 years ago.
There's a sentiment that data center buildouts are priced into the risk of rapidly depreciating GPUs.
But the models want to learn. Token prices are falling so fast that you can get more intelligence out of GPT-5.4 on a 3-year old H100 than GPT-4 on a brand new H100.
@dylan522p @dwarkeshpatel
English
