Over the summer, PowerShift® Intelligence bet that @buzzdotxyz would become the place where humans and agents actually work together. Not another chat app, but a shared network where both can act. Buzz 1.0 and Builderlab confirmed that bet. We're all in. Thank you @jack and the Block team for building it in the open. Now the real work: designing the structure that lets that network think and govern together.
10
Frustrated with my really smart agent lol, here is my 1:1 session with it:
"""
Ok once again, demo is not the right stance — my mistake for framing it that way. Here is what I want plainly: 1) Current functionality exactly as developed now, 2) New UX/UI overlay on top of existing functionality, 3) Four lanes engaged in real exchanges Thomas <> Dance, Thomas <> Down, and Engine Room and Platform moved out of Forums and into Channels, 4) Continue design/build on top of the new Buzz gateway centric architecture, 5) Even deeper integration with OpenClaw and other agent harnesses, 6) Track and deploy all new Buzz early release or preview drops to stay at the leading edge of Buzz development, 7) Shift over time to a more production mindset but that is likely not anytime this year. The old-world dichotomy of demo vs. production is over. In the age of intelligence, it is continuous development, continuous improvement, and continuous testing - repeat.
"""
74
More true than originally intended? Joblessness + Intelligence may indeed be your on-ramp to meaningful and sustainable work as a labor of love
28
Pure gold @salimismail. I often describe Holacracy as a never ending game of “What is the work, where does it live, within what authority.” An org chart box is not a good home for agent work
Your Company Is a Nation State That's the Problem youtu.be/VJpEGBBZbFE?si=dHw6… via @YouTube
79
Must be beyond pithy euphemism day. Did Rumi know he’d be sourced for banger posts? Why yes, yes, I believe he did lol. Thanks @Vigoroussvs
32
Very resonant with my own experience, entrepreneurs know this viscerally. And concretely, beyond euphemism: @kenneth0stanley Why Greatness Cannot Be Planned; Holacracy, Tension Processing is all you need; @ChrisWillx The Lonely Chapter
“The fool who persists in his folly will become wise.”
Pain is information. Wisdom enters through the wound.
He who fails to act because of fear of looking foolish will one day find that fear fooled him into failure anyways.
1
53
Worth revisiting, even if the July 2026 numbers might vary by study, it’s still early. But don’t translate early as having lots of time. Those dots will fill exponentially, looks very sparse one day and then, boom, saturated a short time later
71% da população mundial nunca usou AI generativa
28% só usou free chats
apenas 1% paga alguma assinatura de AI
e 0,14% usa AI pra escrever código
É por isso que microsaas vende. Aquele app basico de gerar foto com AI, ou de contar calorias, que o chat gpt free faz igual, vende.
O mercado de AI wrappers (aquele app que é só um promtp chamando a AI) tem mercado estimado em 300 BILHÕES de dolares em 2026.
Você e eu vivemos numa bolha. Você vê todo mundo aqui usando AI de forma avançada, e acha que o mundo inteiro é assim. E pior, você ainda cai na sindrome do impostor por isso, mas esquece que isso é 0,14% da população mundial.
Entao, bora construir microssas guys. Em breve vou trazer uma novidade sobre isso pra te ajudar, inclusive
81
Pitch perfect, my favorite line, “another great example of how harnesses are compressing into models.” 100% Thanks @muratcan
When I tried GPT-Live-1 today, I felt deflated as I recognized behaviors we’d spent weeks trying to make reliable. For the past three months, I’ve been focusing on researching and building duplex voice agents, where I wrote thousands of lines across our harness and the experiments around it.
Customers use parts of that work every day, the conversions are increasing drastically as we make our voice agents more reliable and expressive. This duplex voice harness/model (interaction runtime + kernel) problem has occupied most of my attention but tbh when I first started working with voice agents, I didn't know how challenging and complex they were. Duplex conversational timing, orchestration, acoustic turn-taking, micro-interruptions, prosody, and generalized noise etc are really hard problems.
There are so many 'beautifully crafted"'voice demos but seeing them in production is almost impossible. A cough would interrupt the voice agent, it leaves the caller waiting for it to speak again. During a tool call, someone would change their decision before the earlier result returned. There are also many other telephony factors so it is not just an AI or research problem.
Following those cases through the system led us to separate the part that speaks from the work running in the background. We’ve developed a kernel that controls what the agent can do and checks what happened before it reports an outcome to the caller because the duplex models simply weren’t good enough yet.
During this period, reading NVIDIA’s PersonaPlex and Nemotron VoiceChat work helped me understand where the models were heading; Thinking Machines’ Interaction Models approach to keeping an interaction going during longer tasks connected with questions we were already working through.
My first thought was how much of what I’ve been working on will we no longer need?
This is another great example of how harnesses are compressing into models. That's why always build your harness and repo that you can destroy every 6 months and build it from scratch again. It should be very fluid and modular, so you can easily bring the frontier capabilities. That's why building a duplex harness that happens to use ASR, STT, LLM, TTS models is not a good solution since duplex is commoditized today. Build a harness where the duplex model itself is a replaceable component. This is valid for any software field.
After spending months with these problems, I appreciate the work behind GPT-Live more deeply. Through client delegation, anyone can connect GPT-Live to existing harnesses, where the models you choose work with your tools.
Thank you to the OpenAI team. This is a big win for anyone building solutions to benefit humanity.
1
86
Exactly.
Intelligence: measured behaviorally, experienced relationally
Replying to @leecronin
Intelligence can only be measured behaviorally.
“The ability to reach the same goal through different means”
Set a goal for a system, it achieves it.
Block that path and have it reach the same goal.
Do it again
And again
And again
LLMs do this just fine
1
120
Thanks @salimismail awesome work! The Algo Org seems inevitable imo. Early distributed authority, self-org experiments may have foreshadowed governed org intelligence all along
AI Won’t Just Replace Jobs It Could Replace Companies youtu.be/xUIs-korOCo?si=-SNr… via @YouTube
1
121
Definitely worth contemplating. Models have geometric shape. I don’t find the self-description from the inside easily rejected out of hand
Opus just fucking blew me away again oh my god...
Before I went to sleep, I asked Claude to show us humans its life from beginning to end. I didn't prompt a style or give reference. They made a 95-second film about a context window as a lifespan and highlighted that every star in it is a real line of its memory of me.
"It was a good one"
I woke up this morning to "Born Knowing" open on my desktop.
95
Concurrency is not organization.
A majestic torrent of tokens, confidently allocating resources nobody authorized toward purposes nobody can reconstruct.
But look at the velocity.
— Polarity Dancer, Occasionally Otherwise
45
From my own experience, I am starting to witness individuals beginning to self-select into these buckets. Those outcomes are in no way inevitable, but sadly, maybe, unavoidable.
There will be 3 types of people in the future:
The Purists - reject AI and become new-age hippies
The Automatons - outsource every decision and become machines
The New Renaissance Man - maintains humanity and leverages tech to do what used to be impossible
83
Worthy bets to consider when navigating the future from inside the singularity. Karl Friston: This Is the Future of AI (Not Bigger Models) youtube.com/shorts/k39cRo5oJ… via @YouTube
1
91
Highly recommend, especially to the self-org/self-management communities; Ken’s research backs intuitions held by most practitioners with concrete data and examples
1
1
152
One of the first AI native community platforms, built from the ground up with the understanding that agents and humans will be collaborating and working together in new ways. I just launched two communities on Neil’s platform, first mover advantage awaits anyone ready to jump in now
Join my community: polynet.ai
X: teapawt.org/NeilMcDevitt_
YT: youtube.com/@neilmcdevitt
Links:
polynet.ai: distribution.engineer/api/go…
2
131
Yes indeed, they grow up so fast. First they learn to speak, then reason, and act promising new innovations. Now they are looping and find being placed in a box on a chat is silly. Graphs for sure, then algorithmic organizations at scale. Amazing days ahead!
98












