Date: 2026-08-27 Attendees: Blake Stewart (note creator), Pierce Freeman (pierce@freeman.vc), Blake Emal (blake@megaphonestudio.com - joined briefly) Granola ID: 91ff899f-ad62-4b0e-ad30-8d8b493cfa05

Summary

Content strategy sync for Pierce Freeman. Pierce built a local trending-topics pipeline that scores momentum/acceleration and surfaces comments to jump on - getting decent results (56, 19 views on best hits; 11 new followers last week). Covered: Bill Gates AI piece, AI-generated viruses research (bacteriophage as drug delivery), Nvidia earnings read, Freeman’s new VM-based workflow orchestration engine (competitor to Temporal/Vercel Workflows), and Apple keynote predictions for rumored Sep 8-9 event. Hot Agent Summer video underperformed (260 impressions, 1 like) - probable cause: no subtitles.

Decisions

  • Pierce sticking with his own comment strategy for now; open to receiving suggested posts from team later.
  • Draft Apple keynote predictions post before Sep 8-9 event, follow up with reaction post after.
  • Blake to audit Pierce’s recent posts; start with Hot Agent Summer video assessment.
  • Blake to delete the duplicate Thursday recurring meeting (stay on Tuesdays only).

Action Items

  • Blake Stewart: Audit Pierce’s recent posts and send feedback. Start with Hot Agent Summer video (subtitles, content patterns).
  • Blake Stewart: Draft Apple keynote predictions post for pre-event publish.
  • Blake Stewart: Delete duplicate Thursday recurring meeting (next sync = Tuesday only).
  • Pierce: Review Grafa drafts Blake queued this week.

Notable Transcript Excerpts

On Pierce’s trending pipeline:

“I came up with this local pipeline, which is my trending topics calibrated to my interest set… a calculation of which ones are gaining momentum or acceleration and then suggesting that to comment on, which has been how most of my last few comments have ended up popping off.”

On Bill Gates AI piece:

“People have been raising flags about this for at least the last 10 years… I kind of welcome Bill to the camp of people that are being concerned. But there’s one quote: ‘I would support a slowdown in AI research if anyone has a good idea of how to do that’… It just goes back to my conversation with Rob - there’s so many people who are like, we would totally support some well thought out policy, but everyone says that, no one’s actually coming to the table with a policy proposal.”

On Nvidia:

“Every conversation I have with people who are in SF right now are pretty much raising money to just spend it on Nvidia chips… The conversation’s like, maybe if we have 90% of compute from Nvidia, can we have 10% from something else that’s cheaper… It’s not like anyone is saying ‘can we just avoid Nvidia entirely.’”

On the VM-based workflow orchestration engine (competitive moat):

“We’re the only one that’s compiling down code… everyone else basically has to boot up another Python interpreter to run these workflows again… We finished a big refactor to switch the underlying workflow runner engine to use virtual machines instead… anytime it’s fully blocked, it can recognize that and save a snapshot of itself… just write that out to a database.”

On trust and missed deadlines:

“The guy that I’m working with is unbelievably talented, probably top engineer I’ve ever worked with. And because of that, we’ve built up a lot of trust capital… I know he’s not yanking my chain… He’s spending nights, weekends making material progress every day.”

On Apple keynotes:

“If Samsung hasn’t cracked the code on foldable displays, Apple won’t either… I think Apple keynotes have kind of lost their sheen… There was such an excitement about something being live and you could feel the hum of the crowd… Live demos created stakes and believability that video can’t replicate.”

Full transcript available in Granola (id: 91ff899f-ad62-4b0e-ad30-8d8b493cfa05)