Chasing hard problems cause life is too easy | prev @AMD @CarnegieMellon

Joined August 2017
12 Photos and videos
if you ever needed a sign to move to nyc, this is it
We hosted SF's only co-ed mixer with more women than men. 45 women, 35 men, and 4 relationships I personally know came out of it. And I'm the guy who went viral for staying "single until series B." Here's the storytime: One random Sunday I tweeted that there should be a singles mixer for YC and a16z speedrun founders. The shitpost hit top news on X. Consensus: "no women will sign up." Instead, dozens of women DMed me: "please make it happen, just keep the ratio right." @dianeemccormack from Fondo offered to run it. @peggy_wang from SparkChat & @parthma said they're single and in. While we planned, I went viral again with my bro @contextconor for joking I'd stay "single until series B." @gregisenberg called it the single worst trend to come out of silicon valley. People called me "everything wrong with SF," said incels were killing the city's dating culture (fun fact: it doesn't exist). So now I was the "single until series B" guy throwing SF's biggest singles mixer. The irony was not lost on me. But a million views, thousands of DMs, and hundreds on the waitlist were counting on me. @pulley signed on to sponsor. So I leaned in. I asked @peggy_wang to be the face of it, dunking on the techbros: sign up so you don't end up like me, single until series B. I was guiding people to a treasure I couldn't possess. And it worked. Best event I've ever hosted. Curated people from every walk of life, not just YC founders, actually hit it off. The girls' group chat is still popping lol. Moral: you can have your cake and eat it too. It's just harder. You can stay "single until series B" and still help everyone else find their person. As @gregisenberg put it in his tweet calling us out: you can have it all. (all love, I like his work) This is independent of my startup @KoyalAI, which helps people make films with AI. Yes, also controversial. I can't help it.
3
1
9
9,071
i bet they will remain single way past their series b
who needs a gf when you can t-pose with the homies Single until Series B - out this summer starring @contextconor and me.
3
8
2,593
congrats to mr musk for being worth over 1 crore crore INR
1
224
better practical effects than oppenheimer - space is hard!
Blue Origin's New Glenn just blew up at LC-36 while attempting to Static Fire ahead of NG-4. nsf.live/spacecoast
292
.@IlluminatedArts cooked w the laser space canons hope this becomes a mainstay
2
182
hosting Technically Single last night in SF was a blast! curating beautiful rooms full of genuinely interesting people and watching connections form is one of the great joys of my life grateful to everyone who came, and to my co-hosts @dianeemccormack, @peggy_wang, and @meh_agarwal DM me if you want in on the next one
Technically Single brought 70 builders together for a night of real conversations, new connections, and a very good time. SF at its best ❤️ Can’t wait for the next one… @KoyalAI x @digipals_app x @Fondocom x @pulley
1
8
428
nice of @Zakirism to be promoting small time creators
6
1,400
at a founder dinner in sf, spoke only about abg cmos and chinese peptides. how do I signal for help?
2
669
> lose trademark wars for the tapisserie dial & the octagonal bezel > slew of temu replicas in the $200-500 range > need to reclaim the market without diluting an icon > controversial lanyard pocket watch at $450 > free marketing for swatch, brand awareness for AP
May 12
Meet the Royal Pop Collection, a groundbreaking union of AP's iconic Royal Oak and Swatch's POP watches powered by a new, hand-wound version of SISTEM51. Available as of May 16, at selected Swatch stores worldwide. #RoyalPop
406
Ladies: if you are looking for a summer project, here is one I found on the streets of SF Go register for @meh_agarwal & @peggy_wang's curated single's mixer
I got confronted for being "single until series b" on the streets of SF yesterday. Without delving into a lot of details, I was with my friend @peggy_wang when he came up with a mic. He looked like a cross between @adilmania and @VirajAla21. Something about a single founders mixer. @contextconor was there too for some reason. Got the footage. Posting before he does. Judge for yourself.
1
1
1
183
Parth Maheshwari retweeted
Replying to @thatssodhawal
born too late to be a polymath. born too early to be a generalist. born just in time to be unemployed
4
3
31
1,008
bringing the most exclusive founders singles event to sf next week sign up for some pure potential play
we solved the hardest challenge in SF: finding love. the most curated mixer ONLY for technically single founders (situationships don't count). i tweeted last month that someone should host one for single yc a16z founders. my dms flooded with female founders panicking about the gender ratio. i am, unfortunately, a man. couldn't fix this alone. then Diane from @tryfondo said: let's actually do this. and @peggy_wang (2x YC founder, female) & @parthma (man 🥲) joined to fix the ratio from the inside. presenting Technically Single: a curated mixer for venture-backed single founders ready to mingle* May 15. 6–9pm. *if you've pledged single until series B, you can still join to pre-schedule dates. rsvp: luma.com/cw9b23l4 world peace starts here.
628
$AMD dropped its Q1 ’26 earnings yesterday, and the stock is trading around $420 (>4x last year’s slump). Some thoughts on the rally as a former @AMD engineer. Not investment advice. I worked on the CPU core team in pre-silicon verification. Most of the job was finding weird bugs before tapeout, validating microarchitectural behavior, and making sure the core could survive the long tail of real workloads. That experience shaped how I think about semiconductor demand. The AI story is not just “more GPUs.” Real data centers are systems. GPUs do the heavy lifting, but CPUs still matter for head nodes, orchestration, networking, memory management, inference serving, agent coordination, and the general-purpose compute around accelerated workloads. As AI shifts from mostly training to large-scale inference, agents, and personalized compute, the architecture gets more hybrid. CPU/GPU balance matters more, not less. That's the part of the AMD thesis I think is still underpriced. EPYC, Instinct, client, embedded, semi-custom: the company has exposure across a lot of places where compute demand is compounding. The Helios rack-scale system, the Meta and OpenAI deals, the Intel collaboration on AI Compute Extensions — none of those happen without a multi-year bet on full-stack compute placed back when the market was pricing a GPU-only world. Working at AMD was a lot of fun. Hard problems, sharp people, and very little distance between the code you write and the physical reality of shipping silicon. Exciting times ahead. @LisaSu
1
208
be michael singer > 20s: becomes aware of, and annoyed by, his inner voice > obsesses over quieting it → starts "the experiment": say yes to whatever life brings, without personal preference interfering > drops out of his Econ PhD to live in the woods and meditate > buys 10 acres in Alachua, FL and builds his shack himself > submits his first book in lieu of a dissertation → starts lecturing at the university > keeps meditating; people start following → founds Temple of the Universe (1975) → 900 acres today, still operating > friend asks him to build a house → "Built with Love, Inc." → builds many → exits > picks up a computer at RadioShack → self-taught programmer → builds custom software > asked to build Medical Manager → 35 years → IPO → WebMD merger at $2B > booked by FBI for securities fraud → 7-year ordeal → fully acquitted > writes books → bestselling author src: The Surrender Experiment. 3/5.
1
109
In the city of tech nerds, one can also choose to be a watch nerd @wornandwound #windupwatchfair
1
2
128
Process is the bottleneck, not chemistry. Howmet and PCC are the TSMC of single-crystal blades. GE/Siemens/MHI have CMSX-4, but they still need to queue for casting slots. New foundry > new alloy.
You cannot buy a new gas turbine until 2030. Order books at GE, Siemens, and Mitsubishi stretch to 2029. Turbine prices have nearly tripled since 2019. Every AI data center needs power and every gas plant needs a turbine. And every turbine has one part that bottlenecks the entire industry: The blade. It has to survive in gas 500°C above the melting point of the metal it's made from and spin at up to 20,000 RPM under 10,000 g of centrifugal force. Each blade is grown as a single crystal of nickel superalloy, pulled through a vacuum furnace at 3 mm per minute. A set of blades costs $600,000 and takes 90 weeks to grow. The same metallurgy powers modern jet engines. Only 3 companies on Earth can build one. China spent $42 billion trying to catch up. They bought a Russian fighter engine, took it apart, and copied every part. Their copy ran 30 hours between overhauls versus 400 for the original. Modern Western engines run 4,000. You can reverse engineer the shape of a turbine blade. You cannot reverse engineer 60 years of metallurgy.
1
103