@mrgadgetstudio Same thing outside code. The work isn't the slow bit, me saying yes is. I've started letting bots make the obvious calls and only bring me the real ones.
@kanavtwt Memory is the bit that breaks first for me. My bots forget a rule I gave them yesterday and confidently redo the thing I said not to. Keen to see if this fixes that
@nvandewetering Depends what you hand them. Mine are useless at 'go be productive' and great at boring stuff with a clear finish line, like filing my bookmarks or prepping tomorrow. Terminal for building, bots for life admin
@brian_millot Ship the onboarding before the features. Nobody sees your best screen if signup loses them first. Let people play before you ask for an email!
@alexhillman Inbox triage, calendar, a morning brief, house hunting, even my wardrobe!
Best one is my reading pile. My X, Substack and Reddit saves get filed into Notion with a line on why each matters. 88 X bookmarks in the last 4 days alone.
Do yours have limits on what they can do?
Here's one outside code. I run 20-odd AI bots on my life admin.
Made this their rule today. Obvious stuff, they decide. Real judgement calls come to me.
My reading bot went first. Filed 6 of today's posts, skipped 2 as too thin, told me after.
A few thoughts about the chief of staff pattern that I haven't seen before, but would love to learn more about:
- 90% of the time, all my agent needs between turns is a ๐ keep going buddy
- the next 9% is multi choice, often with an obvious best choice
- only the last 1% reauires ground up judgement from me
I've designed my Chief of Staff agent to primarily act as a judge BETWEEN each turn depending, deciding which of those 3 buckets is appropriate.
Bucket 1 keeps work moving.
Bucket 2 makes it easy for me to approve (and even automate some of those approvals).
Bucket 3 get escalated to me quickly.
My whole @jfdibot was fundamentally built around agent turns as a building block, which in hindsight turned me into the bottleneck to decide what happens between every turn.
This pattern feels like a magic trick in more ways than I can describe.
Also, as I've gotten better at working with agents (and thanks to lessons I learned working with @mattpocockuk), I've learned to front load the decision work and build in verification layers.
I just don't see this being done often outside of programming work, so I'm genuinely interested in more examples!
@bot I just wanted to pass notes to my Grok Bot, so have setup a shortcut that records my notes, emails it to my bot to action later.
Is this just a note to Grok Bot or a conversation?