dash

making agents dial phones @phoebe_work

[dash @DashellFlynn]

unbelievable alpha for a short

[unusual_whales @unusual_whales]

Bank of America spends $250 million a year on GLP-1 drugs for its employees, CEO has said

[dash @DashellFlynn]

top tier llm managers

[Phoebe @phoebe_work]

happy international intern day!

[dash @DashellFlynn]

@casperdongg very much agree

[dash @DashellFlynn]

home care agencies need ai labs too

[Keegan Wang @WangKeegan]

Excited to introduce Phoebe Garden, our experimental division and blog. 🌱

Building @phoebe_work has led us to hard, subjective problems that don’t fit the paths that the rest of the AI industry is pursuing.

We built The Garden to share the seeds we've planted and the ideas that we're growing.

[dash @DashellFlynn]

incredibly bullish.

[Justin @jrwoodbridge]

I'm excited to announce that @phoebe_work has acquired Alden!

Alden, founded by Catherine Wang in 2024, set out to solve one of home-based care’s biggest challenges: coordination.

Along the way, the Alden team has built a great product and partnered with some of the country’s largest home care providers.

I’m excited to join forces with a team that has taken the same thoughtful approach to the market that we have: earn trust slowly, deliver real ROI, and optimize for the long-term by building a quality, durable product that delivers on its promises.

Thank you to Catherine and the team for trusting us with what you've built. To Alden's customers: We are excited to work with you!

[dash @DashellFlynn]

imagine this on ur cap table

[dash @DashellFlynn]

my @airthings monitor is about to have a field day

[Eric Snitil @EricSnitilWx]

SMOKE ALERT: A heads up, new modeling of active wildfires drives a plume of thick smoke across the Great Lakes over the next 48 hours. Smoke will begin sagging southward from Canada into the United States into Wednesday morning, expanding further into Thursday.

While much of this smoke will exist aloft, there are early indications some of this wildfire smoke will mix down to the surface. This will impact air quality, particularly for those who are sensitive. Aside from air quality issues, there will be a noticeable haze in the air with impacts to our sunrises & sunsets in the form of more vibrant coloring. More to come on any possible air quality alerts as updates arrive.

[Justin @jrwoodbridge]

Wrote a bit about how we are building Phoebe / post-agent companies and living the Factorio dream with @yrechtman

[dash @DashellFlynn]

elite placement

[Taelin @VictorTaelin]

this is my personal singularity moment

this post may sound like a paid ad. I only wish. I'm concerned, more so than happy. the world is changing, and, among the scenarios where AI goes terribly wrong, inequality is the most realistic, yet, the one Anthropic seems to be the least concerned about. I'm glad OpenAI is taking the opposite stance: *personal AGI for everyone*. I think this is a commendable position in the times we live. but who am I in the queue of the bread?

anyway, Fable is here, so I'll just report my first-hour experience

first of all, all my pet prompts are solved.
→ λ-calculus puzzles
→ bug questions
→ one-shot apps
all are trivial to it.

I don't have anything harder other than my
ongoing work

so, in the last several days, I've been toying with HVM5, a new interaction net evaluator with a faster loop.

after writing the first version, I left 32 GPT-5 agents working for ~20 hours each. this resulted in up to 2x speedups, but the file size increased by 2-fold and quality decreased significantly.

I then simplified the whole thing into an even simpler core, and left Opus 4.8 and GPT 5.5 optimizing it for 8 hours. Opus got a legit 6% - 34% speedup in most benches. GPT got better results, but, sadly, an unusable file.

I then asked Fable to optimize it.

2 hours later, it landed a 1770% speedup in one case, 100%+ in other 4, and 22% in average. yes, in 2 hours it outperformed me, opus 4.8 and a swarm of gpt 5.5 agents, by one order of magnitude.

that could not possibly be legit. "it must be hardcoding the benchmarks" (GPT trauma). so I read its explanation and what it did was, indeed, the most high impact optimization one could try first. seems like HVM5 was wasting a lot of time garbage-collecting unused branches of pattern-match nodes. I had optimized that for static mats, but not for dynamic mats. skill issue. Fable figured how to do it for these, resulting in a massive speedup in some benches

but wait, is that *correct*? I'm not sure yet, it is credible, but this is the kind of thing that is very easy to get wrong on interaction nets. the problem is, when I was ready to start auditing Fable's solution so I could tell whether it was buggy or legit, it interrupted me to tell me it had found a massive bug on the code *I* had written.

... wait, what?

so... for garbage collection purposes, I stored a bit on lambda term pointers that meant "the variable bound by this lambda has been freed, so, its lambda must free whatever argument it is applied to". that's fine. yet, on duplicator nodes, I also used the same bit to mean "one of the duplicated variables was freed, so, treat this dup as a passthrough no-op". so, if a lambda entered a duplicator, it would mistake the lambda's collection bit for its own, resulting in corrupted interaction!

that's a mouthful, why I'm writing this?

just so you can appreciate the sheer absurdity of what just happened. I didn't ask it to find bugs. I asked it for an optimization. and even if I did ask it to find bugs, this bug is so astonishingly subtle and specific, identifying it takes mastering the domain to an extent that it beyond even me. I'd easily need hours or days to fix it, *if* I ever came across it. chances are it would just go unnoticed. and Fable found it and fixed it like it was nothing, while it was busy adding a 17x speedup to a file that neither I, nor Opus 4.8, nor a fleet of GPT 5.5 managed to barely make 2x faster.

oh and there is also another tab where it is also ripping through Bend's codebase and finishing everything I had to do

I don't know what to say anymore

this isn't about Anthropic or OpenAI, this is about our collective future as a species. the world is changing, and we need to be aware of it, and discuss how to handle this change.

receipt below . . .

[dash @DashellFlynn]

bro how many receptors they gonna keep agonizing

[Doğal Yasalar @dogalyasalar]

Eli Lilly "2839-LB" adında 5 reseptörü aynı anda aktive eden ilacını duyurdu. (zayıflama iğnesi)

Bu ilaç:
— GLP-1
— GIP
— Glukagon
— Amylin
— Kalsitonin
hedefliyor.

Retatrutide’dan daha fazla kilo verdiriyor (obez ratlarda)

Daha önce zayıflama iğnelerinde:
— Semaglutide → iştah azaltımı
— Tirzepatide → iştah + insülin optimizasyonu
— Retatrutide → iştah azaltımı + yağ yakımı artışı

Retatrutide zaten güçlü çünkü:
— glukagon → enerji harcaması artışı
— GLP-1 → iştah azalması

Ama bazı eksikleri mevcut:
— leptin sinyali zayıf
— erken doygunluk sınırlı
— yeme davranışı tam baskılanmıyor

2839-LB’nin avantajı:

1) Amylin ekleniyor
- beyin yeterince yedim sinyalini çok daha erken veriyor
- binge eating ciddi düşer

2) Kalsitonin reseptör aktivitesi
- amylin etkisini güçlendirir
- tokluk sinyali stabilize olur

3) 5 eksen aynı anda:
- iştah ↓↓↓
- gastrik boşalma ↓
- enerji harcaması ↑
- insülin kontrolü ↑
- nöroregülasyon ↑

artık ilaç firmaları obezite = sadece iştah problemi fikrini terk etti. obezite = çok eksenli hormonal network bozukluğu olduğuna kanaat getirdiler. 3–5 yıl içinde insan denemelerine girerse şaşırmam ama piyasaya çıkması en az 7–10 yıl.

[dash @DashellFlynn]

@eglyman just buy this and put it on ramp travel, game won.

[Alex Slater @aslater]

Stop wasting time.

Speed booking comparison of Soar vs Google Flights:

[dash @DashellFlynn]

@jameygannon @eightsleep is it still avail? in manhattan can coordinate pickup

[dash @DashellFlynn]

my mom just sent me the @WeAreLegora campaign, she doesn’t even know what Legora does, i call that success.

[dash @DashellFlynn]

monitorsituationmaxxxxing

[Massimo @Rainmaker1973]

Inside the BBC News Office: the monitor set-up is wild

[dash @DashellFlynn]

watch figma buy these guys in 0.8 seconds to catch up

[Higgsfield AI 🧩 @higgsfield_ai]

THE FIRST EVER AI motion design tool with full real-time control.

Higgsfield Vibe-Motion is LIVE.
Motion Design was never that easy.

One prompt creates any motion design you can imagine. Refine every parameter live on canvas. No creative limits.

[dash @DashellFlynn]

totally agree, have interviewed thousands and was baffling to me how many people answered “pass”
to legit interview questions

[Zane Hengsperger @zanehengsperger]

I've interviewed hundreds of people in the last 5 months and hired 7 (hiring more now)

the bar is on the ground, it's so easy to be 99% better than everyone

If you want to stand out in a job application process, here's three stupidly easy things you should do immediately:

[dash @DashellFlynn]

@AjcheGustavo

[dash @DashellFlynn]

@theamelia__ don't sell that bitcoin

[dash @DashellFlynn]

@andruyeung bridges nyc.