AI as Distributed Team Cognition
Key Points
- The NASA space‑shuttle story illustrates that critical expertise often resides in the collective interactions of a team, not in any single individual’s knowledge or documentation.
- Current discussions about AI focus heavily on individual productivity hacks, overlooking how AI fundamentally reshapes team dynamics and collective cognition.
- High‑performing product teams that thrive with AI treat it as a distributed teammate, establishing shared rituals, prompt libraries, evaluation norms, and letting AI handle coordination tasks that once required meetings.
- Most organizations, however, simply drop AI into existing workflows without redesigning team practices, risking the loss of the very collaborative knowledge that enables breakthroughs.
Sections
- Losing Collective Knowledge to AI - The speaker warns that AI hype’s focus on individual productivity risks eroding the vital, tacit team-based expertise that made feats like the space shuttle possible.
- Team AI Needs Shared Context - The speaker emphasizes that effective AI adoption depends on embedding it into collaborative workflows by deliberately curating and feeding shared context, rather than relying on isolated individual prompts.
- AI as Collective Team Intelligence - The speaker urges teams to reflect on whether they use AI simply to speed up existing habits or to cultivate true collective intelligence, and invites discussion while referencing a detailed Substack on performance signals.
Full Transcript
# AI as Distributed Team Cognition **Source:** [https://www.youtube.com/watch?v=dohW_oogbis](https://www.youtube.com/watch?v=dohW_oogbis) **Duration:** 00:07:37 ## Summary - The NASA space‑shuttle story illustrates that critical expertise often resides in the collective interactions of a team, not in any single individual’s knowledge or documentation. - Current discussions about AI focus heavily on individual productivity hacks, overlooking how AI fundamentally reshapes team dynamics and collective cognition. - High‑performing product teams that thrive with AI treat it as a distributed teammate, establishing shared rituals, prompt libraries, evaluation norms, and letting AI handle coordination tasks that once required meetings. - Most organizations, however, simply drop AI into existing workflows without redesigning team practices, risking the loss of the very collaborative knowledge that enables breakthroughs. ## Sections - [00:00:00](https://www.youtube.com/watch?v=dohW_oogbis&t=0s) **Losing Collective Knowledge to AI** - The speaker warns that AI hype’s focus on individual productivity risks eroding the vital, tacit team-based expertise that made feats like the space shuttle possible. - [00:03:26](https://www.youtube.com/watch?v=dohW_oogbis&t=206s) **Team AI Needs Shared Context** - The speaker emphasizes that effective AI adoption depends on embedding it into collaborative workflows by deliberately curating and feeding shared context, rather than relying on isolated individual prompts. - [00:06:48](https://www.youtube.com/watch?v=dohW_oogbis&t=408s) **AI as Collective Team Intelligence** - The speaker urges teams to reflect on whether they use AI simply to speed up existing habits or to cultivate true collective intelligence, and invites discussion while referencing a detailed Substack on performance signals. ## Full Transcript
Hey guys, I want to talk about something
that is keeping me up at night that
almost nobody is talking about. It's
about teams and AI and a pattern that's
unfolding in the way we talk about AI
that is prejuditial, that is
anti-team. But first, I'm going to tell
you about the space shuttle. So, to cut
a long story short, NASA forgot how to
build space shuttles.
not in like a vague way where they could
go and get it out of the library. They
actually lost the knowledge. The teams
that built the space shuttle disbanded,
documentation got scattered, expertise
walked out the door. And what's
compelling is that expertise didn't live
in any individual's head. You could get
the blueprints, you could get the
documents, you could get the
specifications, but you still can't
build one today.
The real knowledge about how to build a
space shuttle existed in between in the
collective knowledge of the team. And
that's how all of our greatest
achievements have worked have worked.
It's not in any one person's head. It
lives in the connections between
hundreds of teams in thousands of small
decisions. I've worked at big companies
and I've seen amazing things get built
when teams can work like this.
And what matters right now is that we're
at risk of missing something equally
profound with the way AI is impacting
our
teams. Everyone's obsessed over AI
productivity, prompts, workflows, 10x
your output. Look, I've written some of
that stuff and I am a big believer in
individual productivity gains for AI.
It's measurable, but I think there is a
reason why it is difficult to talk about
team productivity and AI right now. And
I think we're missing something really
big. Fundamentally, we are letting AI
into existing team structures and just
expecting it to work. That's not how it
works. I've been watching teams work
with AI for a long time now, and I'm
seeing a really fascinating divide
emerge. The high-erforming product teams
that I have
seen are doing something completely
different. They're not just using AI
individually. They're fundamentally
distributing their
cognition from the heads that are on the
team now that are human to the AI as
well. That means they've developed new
team practices. They have specific
rituals around AI generated content.
They have common and collective
understanding of how AI works and how
prompts work on their team and what
prompts are working well and not. They
understand eval as a team. They
understand how the workflows that they
did
before in some ways don't need to be
done anymore. They allow AI to pick up
coordination load that would have
previously required a meeting. They are
rethinking from the ground up how their
team works with the assumption that
instead of the team knowledge, you know
how I talked about the space shuttle,
instead of the team knowledge living
just between human heads, now AI is in
the mix too. and you have to find new
ways of working to allow AI to be a part
of that distributed
cognition. On the other hand, most teams
don't work that way right now. Most
teams get a chat GPT subscription.
They're happy to generate ideas
individually. They often take ideas
uncritically. They may inappropriately
substitute a chat GPT chat for product
requirements. I've seen that done. The
difference isn't in the tools. It's not
in whether or not you got Chad GPT or
Grock or Gemini. No, it's in how the
tools reinforce team culture. Look,
we're a very individualized culture on
the internet. I get why we get excited
about individualized productivity gains
with AI. It is a big deal. But teams
build amazing things. And I think that
we are missing a revolution in how teams
function.
For the first time in human history, our
intelligence isn't just living in human
heads anymore. Parts of our thinking are
starting to live outside our minds, our
decision-making, our problem solving,
our creativity. They're living in the
back and forth interactions that we have
with AI. That's what I'm calling
distributed cognition. And teams need to
figure out how to do this
collectively. Teams need to figure out
how to manage shared context explicitly.
A team needs to not just have a prompt
bible or documentation. Teams need to
think about context as something they
are actively feeding to AI. Key
decisions, refined
outputs, a wide range of inputs that
they deliberately
curate. It needs to feel like a part of
the natural workflow. And I know that it
won't at first, but it's really really
critical to enabling AI to be a
successful partner on the team and to
actually harvesting not just individual
productivity gains, but team level
benefits. Because if the individual
comes and says, "Hey, I wrote that
product requirements." And they feel
happy and they give it to you and it's
not good, but they got it done 10 times
faster. The team as a whole did not
benefit. It was just the individual
feeling like they' done a better job.
And I'm seeing that happen too often.
I'm seeing teams forget that they need
to rethink how they make decisions
because now AI is 10xing or 100xing our
optionality. If you can get 10
iterations of a marketing message, you
should probably rethink how you do your
process. But we're not doing
that. And so my sense is the gap between
these approaches is getting wider all
the time. AI models are continuing to
grow. They're continuing to become more
powerful. They're supercharging the few
teams that know how to use them well.
They're offering real benefits to
individuals, but by and large, teams are
being left behind. And I think teams are
a fundamental unit of work. I opened
with the story of the space shuttle for
a reason. Making things is hard. Humans
have figured out that making things
together allows us to do more than we
could do individually. And we have to
figure out how to reformat the team so
that AI can be a fully functioning team
member. Otherwise, we're going to
consistently underestimate the real
potential of AI. Not because we don't
understand the intelligence AI brings to
the table, but because we don't
understand how to make that intelligence
social. The lesson here isn't just like
document everything in markdown. It's
about understanding that intelligence
operates in the spaces between people.
And now we have to think about
intelligence operating in the spaces
between people and also AI. Ultimately,
that's what's going to help teams be
more productive. And so my challenge to
you is basically how is your team
thinking about AI collectively? Are you
just using AI to speed up old patterns?
Are you actually engaging with it as a
form of collective intelligence? I'm
really curious to hear about this. So,
drop a comment if you want to read more.
I actually wrote up um a long substack
on this that talks through like eight
different signals of high performing
teams, signals of low performing teams,
how you break through. Lots of good
stuff in there. And if you don't want to
read it, I don't care. That's fine. Um
yeah, tell me how you are getting to
collective intelligence with your AI
usage on your teams. Cheers.