inguz ᛜ outcomes

still in it

[inguz ᛜ outcomes @hughpyle]

8-hour sprints seem to be working well. 6 would be worth trying too.

[inguz ᛜ outcomes @hughpyle]

Developers: how long should a sprint be? Is 8 hours too long?

[inguz ᛜ outcomes @hughpyle]

@smiling_for_fun No particular recs (maybe Banks “The Player of Games”?) but have experienced both & very happy to chat, DMs open

[inguz ᛜ outcomes @hughpyle]

@threepointone It can work. I built a LambdaMOO out of distributed virtual machine DOs and it almost did https://t.co/hgbdOe7psU

[inguz ᛜ outcomes @hughpyle]

@grove_research Congrats. It'll be very interesting to see where you take this.

If you want to add "more structure" on atproto, see if `atseq` pushes your buttons: https://t.co/Czsne5fBJv

[inguz ᛜ outcomes @hughpyle]

I mean

[inguz ᛜ outcomes @hughpyle]

My new favorite prompt is: “Let’s simulate: …” and some crazy idea.

[inguz ᛜ outcomes @hughpyle]

@KennethFolk Fillmore at Tanglewood!

[inguz ᛜ outcomes @hughpyle]

entire AGENTS.md now: “communicate in E-Prime plain language”

[Geoffrey Litt @geoffreylitt]

in my experience, agent skills can totally remain relevant over time / across model generations, *as long as they are more spec than step-by-step instructions*

eg: specifying a good format for a doc output, a good flow for interviewing a human

[inguz ᛜ outcomes @hughpyle]

is there a costco version?

[ADONIS @adonispara]

The final frontier of consumerism has arrived in South Korea.

[inguz ᛜ outcomes @hughpyle]

@yishan “siiiiithe”

[inguz ᛜ outcomes @hughpyle]

@Mentioum @RhysSullivan 57 is more stealth

[inguz ᛜ outcomes @hughpyle]

"When people communicate face to face, they externalize their models so they can be sure they are talking about the same thing. Even such a simple externalized model as a flow diagram or an outline... serves as a focus for discussion." https://t.co/7c3graf5qa

[inguz ᛜ outcomes @hughpyle]

`dap` is the latest in this series, and I think it has the potential to be interesting:
https://t.co/c32hjlDfZM

[inguz ᛜ outcomes @hughpyle]

I'm continuing to build out "fold over a sequenced log" experiments, because they're a really nice foundation for applications where the users are people and agents.

Here are four infrastructure variants so far.

[inguz ᛜ outcomes @hughpyle]

@noahrshinn “situational apps” https://t.co/vCWCgC8BZq

[inguz ᛜ outcomes @hughpyle]

And this category of "situational apps": I think it's a natural evolution of https://t.co/Qwxb7qAn6z.

Unlike "multi-sided platforms" (ebay, uber, doordash...) where the situation is repeatable: so many of our daily interactions are multi-sided but *unique*. Agents can help!

[inguz ᛜ outcomes @hughpyle]

@RobertHaisfield Iain M. Banks movies would be a good thing right now

[inguz ᛜ outcomes @hughpyle]

@JoshARosen Did you look at the cloudflare “gatekeeper” model? An interesting spin.

[inguz ᛜ outcomes @hughpyle]

And this category of "situational apps": I think it's a natural evolution of https://t.co/Qwxb7qAn6z.

Unlike "multi-sided platforms" (ebay, uber, doordash...) where the situation is repeatable: so many of our daily interactions are multi-sided but *unique*. Agents can help!

[inguz ᛜ outcomes @hughpyle]

I'm continuing to build out "fold over a sequenced log" experiments, because they're a really nice foundation for applications where the users are people and agents.

Here are four infrastructure variants so far.

[inguz ᛜ outcomes @hughpyle]

The sequencer isn't "global". Next experiment may incorporate the interesting "CTEG" work by Simon Foldvik - I think by tracking causal dependencies (as gitseq does) we can efficiently implement cross-partition transactions.
https://t.co/0Ym6AuHGo1