2026-05-31

Fwiw I suspect that the need to keep this voting bloc proportionally larger than any other is one of the major reasons why it is so hard to build new housing here unless it is earmarked for people who will need lots of social services
mattparlmer 🪐 🌷: There are probably ~120000 ppl in SF whose primary income is supported by tax receipts (excluding pure welfare benefits recipients)
If you assume that households which contain one such person vote in their economic interests, the “graft bloc” may be half of SF municipal ballots

Heads up: creatives are starting to vibe code, and it is awesome.
Two things are happening:
1. Creators are starting to find tools like Replit.
2. Replit is becoming more creator-friendly and native to how they work. Example: Canvas.
I'm really excited about what these folks bring. When you get these tools in the hands of creative people, you see magical things happen.
Tomorrow I'm demoing at a second creator hackathon and I'm sure it's going to be fire.

I agree with @JTLonsdale on these suggestions.
I’ll add one more. aggressively expand the most successful public-private partnership in American history: the research university system.
We should 10x the budgets of the NIH and NSF and fund massive dataset buildouts through universities that can be used to power the next gen AI systems. Training on those datasets can be restricted to American entities for a number of years to bolster our industries.
and yes we can reform the parts of academia that need fixing in parallel. There is no world in which America wins in which the majority of new basic science is happening abroad.
Joe Lonsdale: 1/ US biotech is in crisis, right before AI should be saving millions.
China is stealing away our industry and has surpassed the US in blockbuster pharma deals.
The next FDA Commissioner must be a fighter, and have a plan to overhaul the agency, beat China, and unleash cures.

Because San Francisco refuses to build housing and so most service workers have to travel up to two hours each night back into exurban exile in e.g. Solano County and Contra Costa County.
Scott Stevenson: Why does everything in San Francisco close so early? Coffee shops, gyms, corner stores.
I have never experienced anything like it—is surprisingly inconvenient!

Five million users would agree. Resetting the limits tomorrow morning to celebrate.
Time to go /fast
Siqi Chen: nothing like switching to claude for a few days to try out a new model and going back to codex xhigh to remind you how much better 5.5 is right now
it's really not close

卧槽！你随便扔一首 YouTube 歌，它就能直接拆成 6 条独立音轨——人声、鼓、贝斯、吉他、钢琴、其他……全给你扒出来！
这工具叫 StemDeck，GitHub 开源项目，本地跑全程不上传、不注册、不花钱！它基于 Demucs 模型，拆完后给你一个接近 DAW 的界面：
静音、独奏、拉推子、缩波形、设循环、单轨下载、混音导出……想怎么玩就怎么玩。硬核亮点：
本地全速运行：自动识别 GPU，Apple Silicon 上快得飞起
智能分析：自动算 BPM、调性、响度，扒谱党狂喜
超灵活：想只拆几轨、留原声道做参考也行，处理中途随时取消 练歌、扒谱、做 remix、 stem 练习……全都能干。这波真香啊，比那些动不动就收费的云端分离服务爽多了！
https://github.com/thcp/stemdeck自己搭起来玩，数据安全又免费，强烈推荐！

The Magnificent 7 has an imposter.

Hot take: obsidian is a terrible form factor for an AI 2nd brain.
GBrain is functionally best option, but non-technical users can’t reconcile that they can’t “see” what the AI knows.
Very soon, something like @nessielabs will become the top 2nd brain option.

"It's terrifying launching something. I've launched 100s of products, and every time it's like...what if zero people care?"
Here's my new episode with @Shpigford, a solo builder who sold Baremetrics for $4M and is now building 5 products at the same time with AI agents.
We covered the AI skills that he uses to ship:
✅ /build skill to research, plan, track, and build new features
✅ /adversarial-code-review skill to have GPT review Claude's work
✅ /but-for-real skill to get AI to fix its own mistakes
...and more.
Some quotes from Josh:
"The idea of spending months working on something before you put it out for other people to use, I think that's a really bad idea."
"I'm textbook ADHD. My brain naturally is just ping-ponging all over the place. Now I can feed the beast a little bit more."
"My advice is to just fail a lot, because the only way that you'll ever figure out what not to do is by doing the thing incorrectly."
📌 Watch now: https://youtu.be/GdxLaeyu33c
Thanks to our sponsors:
@WisprFlow: Don't type, just speak https://ref.wisprflow.ai/peteryang
@linear: The AI agent platform for modern teams https://linear.app/partners/behind-the-craft

This has quietly been a miracle month in medicine.
In the last 5 weeks we’ve got news on:
- retatrutide, the triple agonist GLP-1 from Lilly, basically melting fat and body-wide inflammation at record levels
- RevMed’s new pancreatic cancer drug showing unprecedented abilities to extend life
- small trial of a one-and-done PCSK9 gene editing therapy for slashing LDL cholesterol
- Mayo’s AI-assisted radiology showing vastly improved cancer detection
- this new therapy for metastatic solid tumors
This stuff is at varying levels of evidence. Retatrutide is ~100% on its way, other stuff needs more clinical trial data. But put it together and we’re maybe on the verge of majorly reducing the mortality of heart disease and cancer, the two leading causes of death in America.
Crémieux: This is actually insane.
97% of people taking the standard of care for metastatic solid tumor got worse by seven years.
But with lorlatinib, that number was only 45% in the same time!
This is an ENORMOUS jump in the quality of cancer care.

Timeline of the Iran War:
🔸2015: JCPOA agreed to. Iran gives up most of its enriched uranium, leaving it with a limited amount enriched to under 4%.
🔸2016-2018: Iran is complying with the agreement.
🔸2015-2025 US intelligence says Iran is not working on a nuclear weapon.
🔸May 2018: Trump withdraws from the JCPOA and begins imposing harsh sanctions on Iran. Iran continues to comply with the agreement for another year.
🔸May 2019: Iran ends its compliance with the agreement, which is no longer recognized by the US and is no longer providing sanctions relief. Although Iran is producing more uranium, it is still at low levels of around 4% for the next couple of years.
🔸Jul 2020: Iran's Natanz facility is sabotaged.
🔸Nov 2020: The father of Iran's nuclear program, Mohsen Fakhrizadeh, is assassinated.
🔸Jan 2021: Iran begins enriching uranium up to 20%.
🔸Apr 2021: While talks were underway to restore the JCPOA, the Natanz facility was attacked again. Iran announces 60% enrichment in response. The 60% enrichment volume is low. Iran says it is a demonstration of capability and is reversible.
🔸Nov 2024: IAEA enacts a censure resolution against Iran
🔸Dec 2024: In response, Iran activates idle centrifuges and dramatically increases enrichment production.
🔸2025: US intelligence threat assessment says Iran is still not working on a nuclear weapon and warns that conflict with Iran would lead to a closure of the Strait, attacks on US bases, and attacks in the region.
🔸Jun 2025: US bombs Iran's nuclear facilities. Trump declares Iran's nuclear program obliterated.
🔸Feb 2026: Iran and the US are involved in negotiations about Iran's nuclear program. Iran reportedly is flexible on its highly enriched uranium, including possible transfer, exchange, or downblending.
🔸Feb 28: During negotiations, the US and Israel launch attacks on Iran. Iranian leadership is killed, military facilities are struck, and an elementary school full of children is bombed.
Iran retaliates by closing the Strait, attacking US bases, and targeting other Gulf nations.
Throughout the war, the US has had 15 bases heavily damaged, 42 aircraft lost or damaged, 13 service members killed, and hundreds more wounded.
🔸Mar 1: After telling journalists several different timelines, Trump settles on 4-5 weeks for the conflict. The administration admits there was no imminent threat of attack from Iran.
🔸Mar 3: AAA shows the national average for a gallon of gas is above $3.00
🔸Mar 7, 11, 15, 20, 24, 26, 27 Apr 7, 8: The administration says the war is won.
🔸Mar 9, 11, 13, 16, 17, 20, 23, 24, 29 Apr 1, 6, 16, 17, 20, the administration says the war will be over soon.
🔸Mar 21: Trump gives Iran a 48-hour ultimatum
🔸Mar 23: Trump extends the time limit for the ultimatum by 5 days
🔸Mar 26: Trump extends the ultimatum deadline to Apr 6
🔸Mar 30: AAA shows gas above $4.00 a gallon
🔸Apr 5: Trump pushes back the time limit to Apr 7
🔸Apr 7: Trump threatens to destroy Iran’s entire civilization if they don't comply, then announces a two-week ceasefire instead of attacking
🔸Apr 21: As the two-week ceasefire is about to expire with no deal, Trump says it is now an indefinite ceasefire
🔸May 1: War reaches 60 days and is now in violation of the War Powers Resolution.
🔸May 7: Despite a ceasefire in effect, Iran and US forces exchange fire. Trump says the ceasefire is still in effect
🔸May 12: CPI report shows inflation reaches 3.8%. Inflation is now outpacing wage gains. Trump says he is not thinking about Americans' financial situation
🔸May 23-24: Administration suggests a deal will be reached this weekend. No deal happens.
🔸Today: No deal has been reached. Gas is $4.34, inflation is still rising, the Strait is still closed, and Iran is still not working on a nuclear weapon

Trump’s “little excursion to Iran” is "universally perceived as a defeat. Almost regardless of the outcome – most likely a return to the…status quo – the war looks ill-conceived, a monument to confused objectives, bad planning and misplaced assumptions." https://trib.al/hR5GtjE

“A UC San Diego report of soaring math unpreparedness is fueling faculty warnings that reliable testing is needed for admissions.
Critics call the SAT inequitable and say high school grades are a good predictor of college success.” 🤦🏻‍♀️ https://www.latimes.com/california/story/2026-05-27/uc-math-professors-demand-return-of-sat-for-stem-admissions

The events of the last 6 months in technology are arguable amongst the most important in human history
The tools now increasingly exist for recursive self improvement of models & agents
We are likely in very early lift off & exponential
Largely unnoticed outside of tech

California doesn’t build so housing will only go up, forever, until more houses are built. All commentary about rents etc is downstream of this.
Norgard: What happens to the SoCal housing market after the SpaceX IPO?
The communities closest to Hawthorne see the first wave. Manhattan, Hermosa, Redondo, and PV get bid up 25% or more. But those markets are already supply-constrained. Today, there are only 39 properties for sale in

Re Hey Garry, GBrain already covers more than half of the lift for this already!
There are 2 main components:
1. Memory indexing (embedding/enrichment) during ingestion
2. Accurate, Low latency, cost-effective Retrieval using the enriched embeddings
I measured GBrain across industry benchmarks LoCoMo, LongMemEval, and the “hardest” one Memory Agent Bench (MAB) — did you know your system crushes some of the leading Labs out there on MAB?
Happy to share the data any time. Invested like $500 to run your system on those benchmarks and get the data because I was genuinely curious. Nice work ser!
DM if you’d like and I’ll send the benchmark data right on over.

My favorite AI skill from @Shpigford is this /but-for-real skill that makes AI double check its work. It includes classic lines like:
"You just mass-produced a pile of changes with the unearned confidence of a junior dev who's never had a production incident.
"Spoiler: you have production incidents *constantly*. The user just doesn't call them that because they're too polite. They call it "can you try again?" which is code for "you failed and I'm being nice about it."
📌 Watch Josh talk about this skill here: https://youtu.be/GdxLaeyu33c?si=dTZK0KPNc_i7wQaf&t=1015
Peter Yang: "It's terrifying launching something. I've launched 100s of products, and every time it's like...what if zero people care?"
Here's my new episode with @Shpigford, a solo builder who sold Baremetrics for $4M and is now building 5 products at the same time with AI agents.
We

Your startup's growth rate is also its learning rate.
Setting an ambitious growth goal and trying to hit it is the fastest way to find out what's broken and how to fix it.
Not just the obvious stuff. Wrong customer, wrong problem, product nobody wants. It also stresses the founding team. You find out fast whether you can handle hard problems under pressure.
Bottlenecks only show up when you push.
If you're not having uncomfortable discoveries early, you're not pushing hard enough.

Everyone knows Pilots vibe with Replit @PeterDiamandis @amasad @Replit
james hawkins: absolutely wild, founders could learn a thing or two from this