Manu Narayan
speaker
373 appearances
1 recordings
1 series
first heard Jul 2026
last heard 22 Jul
Manu Narayan’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 1 in all, peaking in Jul 2026 with 1.
Appearances
You hit on something else which was really interesting, which is
How does this then become more cross-domain as you start to think about the people team learning from the AI from the IT group and learning from others?
And I think we see that really coming about a lot as you think about things like if you're leveraging Claude and you're creating skills that help with internal research, the specific things that you're researching may be different, but the approaches that you take can oftentimes be really similar.
And so we find that we have a lot of skill sharing.
project sharing happening between groups where they can really share that ideation and learn from one another on how they scale out what they're doing.
I think and maybe start with that last one first, if you don't mind.
I think it's a k it's a good question 'cause the there aren't definitions for these things really, right?
What is an agent and how's that different some from some of the other things that we do?
You leveraging a I tend to personally draw a distinction between what's
human invoked versus autonomous.
And when I think about true agentic workloads, I'm thinking about things that are maybe triggered by events that run autonomously, polling and monitoring for things, versus oftentimes skills are a bit more human and person invoked.
It gets a little blurry if you think about things like cowork and other things that can then execute agents on skills on a schedule.
But I think that to me is one of the big differentiators is a skill often is
A skill that's enhancing what a human does and agent something that while you may have a human in the loop as part of the process, ultimately is running a bit autonomously.
As we think about skills internally, I think that's one of the key ways that we actually scale out what we're trying to achieve.
If we take that example we talked about where there's the haves and the have nots of AI usage, people that have figured out how to use AI really well for their job, whether they're using skills or not, they're ultimately creating usually a set of reusable prompts.
Maybe they're using MCP and other things to pull in data and context, even take access.
Yeah.
And the skill helps codify that, which is really important.
But moreover, it actually lets us scale that out really easily to other people in similar roles.
Showing 61–80 of 373 · page 4 of 19
← Previous
Next →