harsh

1.3K posts

harsh

harsh

@hpatel

i build products and companies that solve problems for people. right now, turning AI tokens into shareholder value at https://t.co/JaVhi3IcrN

St. Louis Присоединился Şubat 2011
708 Подписки4.2K Подписчики
harsh
harsh@hpatel·
@thsottiaux i use codex for back end and claude for front end. please improve front end, i would love to go codex exclusively.
English
0
0
1
127
Tibo
Tibo@thsottiaux·
Hello builders. What are we getting wrong with Codex, what can we improve?
English
2.2K
59
2.5K
256.5K
harsh
harsh@hpatel·
woah, can't believe this actually happened! just got @hpatel from the X Handle Marketplace (handles.x.com)
English
2
0
2
584
harsh
harsh@hpatel·
i feel seen tl;dr: my job at teslamr.com since Oct of 2025 (also nice to be founder so you can make 100% of the calls, at AI-speed)
Aaron Levie@levie

The more enterprises I talk to about AI agent transformation, the more it’s clear that there is going to be a new type of role in most enterprises going forward. The job is to be the agent deployer and manager in teams. Here’s the rough JD: This person will need to figure out what are the highest leverage set of workflows on a team are (either existing or new ones) where agents can actually drive significantly more value for the team and company. In general, it’s going to be in areas where if you threw compute (in the form of agents) at a task you could either execute it 100X faster or do it 100X more times than before. Examples would be processing orders of magnitude more leads to hand them off to reps with extra customer signal, automating a contracting review and intake process, streamlining a client onboarding process to reduce as many straps as possible, setting up knowledge bases than the whole company taps into, and so on. This person’s job is to figure out what the future state workflow needs to look like to drive this new form of automation, and how to connect up the various existing or new systems in such a way that this can be fulfilled. The gnarly part of the work is mapping structured and unstructured data flows, figuring out the ideal workflow, getting the agent the context it needs to do the work properly, figuring out where the human interfaces with the agent and at what steps, manages evals and reviews after any major model or data change, and runs and manages the agents on an ongoing basis tracking KPIs, and so on. The person must be good at mapping the process and understanding where the value could be unlocked and be relatively technical, and has full autonomy to connect up business systems and drive automation. This means they’re comfortable with skills, MCP, CLIs, and so on, and the company believes it’s safe for them to do so. But also great operationally and at business. It may be an existing person repositioned, or a totally net new person in the company. There will likely need to be one or more of these people on every team, so it’s not a centralized role per se. It may rile up into IT or an AI team, or live in the function and just have checkpoints with a central function. This would also be a fantastic job for next gen hires who are leaning into AI, and are technical, to be able to go into. And for anyone concerned about engineers in the future, this will be an obvious area for these skills as well.

English
0
0
1
307
harsh
harsh@hpatel·
@nevmed hahah i still do that on my HW3 (99.9% of driving is on FSD), but they've stopped shipping these cool new updates to HW3 it seems which is #pain
English
1
0
1
53
Neville Medhora
Neville Medhora@nevmed·
@hpatel Upgrade! Did 4 errands today, literally hit "self drive" out of the parking lots to another place across town, barely paid attention to where I was, then end up parked in new parking lot. Repeat. Soooo nice.
English
1
0
1
268
harsh
harsh@hpatel·
opus sucks at code but killer in openclaw codex killer at code but sucks in openclaw whyyyy... plz fix, @thsottiaux and @steipete
English
1
0
1
149
harsh
harsh@hpatel·
my mac just autocorrected GDP to GPT so that's where i'm at in the ai cycle
English
0
0
3
75
harsh
harsh@hpatel·
@Shpigford successfully juggling extreme velocity across all projects
English
0
0
2
70
Josh Pigford
Josh Pigford@Shpigford·
with AI, do you find yourself juggling more work across *disparate* projects? or just increased velocity on the *same* project?
English
35
2
26
4.8K
harsh
harsh@hpatel·
feels awfully close to agi!
harsh tweet media
English
1
0
3
257
harsh
harsh@hpatel·
@elonmusk can u plz just make the best coding model @xai, take the whole market (switching costs are 0), then keep making imagine videos
English
0
0
2
77
@levelsio
@levelsio@levelsio·
Okay let's see who can reply to this
English
2.5K
17
2.1K
1M
Nat Eliason
Nat Eliason@nateliason·
@rachelgoodlad I have the “request alpha in your city” link always ready to go
English
1
0
3
136
harsh
harsh@hpatel·
@mert under all the vibecoded new launches looking for pmf is a gigantic ocean of already PMF companies automating everything by vibecoding we've automated 80%+ of our entire company at teslamr.com, all via vibecoding/agenting engineering whatever tf u wanna call it
English
0
0
1
69
mert
mert@mert·
so like has anyone actually gotten PMF with anything vibe coded so far or are we just gonna realize this was a huge bubble in a few months
English
298
25
1.4K
258.7K
harsh
harsh@hpatel·
@FurqanR Why not… Nebula?
English
0
0
1
190
Furqan Rydhan
Furqan Rydhan@FurqanR·
We're growing faster than we can ship content. Nebula needs a Dev Rel / Technical Content person who actually understands AI agents and can explain them like a friend. If that's you, DM me.
English
50
17
302
22.1K
harsh
harsh@hpatel·
big milestone for @xai... I've officially added grok 4.20 as the preferred model in production for a workflow over Anthropic, OpenAI and Gemini. little by little then all at once?
English
0
0
1
141
harsh
harsh@hpatel·
woah composer 2 is... fantastic and fast
English
1
0
2
152
Cathryn
Cathryn@cathrynlavery·
@Shaun__Furman sorry i meant 1 team, multiple agents but managed by a main orchestrator. I actually just reduced my team because there were too many agents, and I realized it just wasn't effective.
English
2
0
1
239
Cathryn
Cathryn@cathrynlavery·
🦞 pro-tip for multi-agent teams A big unlock for me with AI agent team is making agents report to my chief of staff, not me. Just added: → blockers go to him first, not my inbox → 9am + 4pm he surfaces what's stuck + one clear ask if he needs me on something → every escalation logs to autonomy-gaps.md Then I know where to fill in the gap so he needs me less and less.
English
12
2
68
5K
harsh
harsh@hpatel·
@ErikTrautman I have a codex 5.4 high instance on at all times in the openclaw folder and that’s made debugging a breeze. I urge you, keep going. My output has 100-1000x’d on all fronts of running a company… product, eng, marketing, content, ads, sales, literally everything.
English
0
0
1
60
erik
erik@ErikTrautman·
I'm very rusty, but the level of debugging I'm having to do with Open Claw just to get it to the starting line on a clean setup makes me skeptical of the tales of its easy successes.
English
5
0
3
588