Michael E. Gruen
@gruen.us
540 followers 180 following 880 posts
Excited to be here. More at michaelgruen.com/now
Posts Media Videos Starter Packs
github.com/charmbracele...

I confess I haven't done much work on this... I just asked Claude Code to evaluated the top 3 search results through subagents and pick one for the project.
Try having it make you a TUI instead of a GUI—it's not half-bad. (Most of my CLI hackery is in golang, but I'm sure there's fancier stuff out there.)
Haven’t gotten a marketing email from them yet, but that I didn’t want to sign up at all is the problem; but, needed to execute a contract.

There’s always a drift when PII floats between systems and queues, which means there’s too may opportunities to say “oopsie, it was a technical error”.

🐂💩
I'd forgive a 3 person company for this. Not a $5Bn market cap company.
To the box.com executive who approved that:

1. the signup TOS require auto-enrollment in marketing e-mails, with
2. no way to unsubscribe in the admin interface, requiring me to have to search to find this article: community.box.com/end-user-for...

This is asshole design. You should feel bad.
When the computer switches to dark mode, perhaps it's time to consider going home.
Sounds like that packing bubble popping app has some competition.
The older I get, the more I care about cable management.
0 tokens remain. /compact
3 down in today’s mini is misspelled.

K?
Looks like noaa cuts are starting to show
Ron is a machine! Saw him live a few months back.
Oh those custom /commands look money

Now to build a fleet of cli tools.
Seems like a lot of work for periodic version bump ;)

0.4 is still current it seems.
`fly deploy`

could cron it ;)
I can't believe I'm saying this, but: can't LLMs replace all IVRs?
Equally, I’ll say that I am sometimes the pusher on my teams to adopt AI. The skepticism is (rightfully) there, given both public failures and hype.

When you can demonstrate a reasonable application of current AI tech, where metrics are improved and pain is reduced, feelings become measured.
I don’t say this often, but I agree with Gartner’s assessment that we’re entering the trough of disillusionment. Most of these projects will continue to fail, and many CTOs (and Chief AI Officers for that matter—remember Chief Social Officers?) will successfully shrug off the corporate hubris.
Executives, in response, are placing bets on highly speculative AI projects that probably won’t work, but it eases the pressure as it’s something they can talk about and demonstrate responsiveness to CEO and board.

The hype cycle is real.

AI is an overloaded term, much spend fueled by FUD.
What I can say publicly is that CEO pressure is real, but most of them come to their senses and approve the right experiments if they trust their tech leadership.

(Not that they always do.)

This pressure on the CEO comes from board and investors who understand AI even less.
The bigger issue is lack of understand what AI is and can do.

I’m all in on using LLMs for codegen on sane, zeitgeist-approved defaults and stubbing content to be later refined and replaced.
Thick rimmed glasses crew
Fly to SUX and I’ll send you some recs ;)