šŸ› ļø @stedi Prev: @Trek10Inc, @goserverless ā˜ļø AWS DevTools Hero. šŸ§” he/him.

Washington, DC
Joined September 2009
okay. hear me out.
1
5
191
LLMs ain't the only thing out here hallucinating.
117
tbh the only reason I want a duo is for @t3dotcodes and I'm not sorry.
132
Clearly I just found AI in 2020 and squandered my lead.
1
112
So everyone is making like faster typescript... can we start having a race for rust? Having faster rust compiles would make me very very happy.
1
1
209
Didn’t have deshittification on my bingo card for the year, but I’m enjoying the heck out of it… being able to profitably manufacture premium hardware without leaning on software to lock in long term revenue will be a superpower.
3
257
We are going to get an unbelievable amount of slopgames, but the neat part is we can iterate on fresh concepts at an unbelievable rate from sources we'd otherwise never have gotten. Sure 99.9999% will be trash, but the market will sort that out for me.
3
323
Having so many problems with github... we're are gonna try the classic "schedule a port my github org to an HA forgejo on my home cluster for tonight" prompt.
1
272
this is the kind of stuff that makes it seem pretty likely the next OS leap will just be optimized for agents to build personal software on demand.
this was so crazy i had to build a native mac UI for it 17,000x faster than Finder github.com/zeusinsight/Finde…
271
The inline visualizations in @t3dotcodes are amazing.
1
247
it is wild how i can tell across multiple products when inference speed is suffering on a core provider. everything just gets so painfully slow.
153
dots is surprisingly good at digging up coupons for me for sites. no more wading through 38 fake links trying to referral hijack.
189
Biggest thing missing for me now in @t3dotcodes is a cluster mode. Load balancing threads at initialization is great, but once something goes out to a node, spawned off threads get stuck on that node instead of being able to fork out & coordinate across all my available compute.
215
My "shed" turned into a mansion (mini 12u rack) in my basement that self-manages / upgrades / heals with a 5gig symmetric line... it now also runs all my github action runners and @t3dotcodes envs.
Every developer needs an AI shed: An always-on machine running on their tailscale network where the majority of their herdr agents are running.
1
1
335
naw, there is very minimal shared understanding of the capabilities. Even in the coding world, many are still copy pasting from chatbots a few times a day instead of going to stackoverflow.
Everyone uses AI by now. Senior leaders are vibecoding apps. They've adopted claude. They're finding workflows to automate. Folks have their own claws. AI capex is keeping the S&P afloat. Corporate America's geared up. It's all very much mainstream. We're not so early anymore.
1
285
every. damn. time.
2
293
Even amongst my very tech-savvy friends, they have no idea what is possible. It’s just a sometimes wrong Google analog. When I describe even the more basic things I use it for, it’s like I’m trying to describe a cell phone when we just got telegraphs.
98% of US households aren't paying for AI yet More charts in State of Markets II: a16z.news/p/state-of-markets…
1
3
446
Goodness gracious, @t3dotcodes orchestrator v2 isn’t just in a different league, it’s playing some other sport entirely. I have a mini-server rack with self-managing t3 envs, and it is paying off immediately to be able to spin off threads to those workers.
1
1
299
with the advent of astra ultrafast, I can now exercise my procrastination with absolutely zero regret.
Global reset landing tomorrow 10am PST for all paid ChatGPT accounts. Apologies for the slow start with GPT-6.1 Sol, it's now back to running at expected speeds after the massive load spike in the first two days.
1
369
Update... sub one-hour torch of the 80% i had remaining without even trying hard. I'm just gonna pretend that setting doesn't exist from now on.
153