Thomas Brady

753 posts

Thomas Brady banner
Thomas Brady

Thomas Brady

@thbrdy

techno-vajrapāṇi || defend the dharma || Building @notice_tools || writing pop-ai dot co

San Francisco Katılım Şubat 2026
833 Takip Edilen77 Takipçiler
Thomas Brady
Thomas Brady@thbrdy·
From what perspective? You can create skills that deal with this on an atomic level and iterate in loops using a handful of examples for the agent to build off of. I'm building writing skills for my agents now to manage drafting big pieces of my work. It's easier than writing from scratch.
English
0
0
0
29
Ethan Mollick
Ethan Mollick@emollick·
I read a few dozen pages of this and it is not bad for LLM fiction, but also very very LLM-y, from the themes to the fact that there are lots of staccato conversations and meaningful silences and overwrought metaphors and very little differentiated character development.
Nous Research@NousResearch

Hermes Agent wrote a novel. "The Second Son of the House of Bells" runs 79,456 words across 19 chapters. The agent built its own pipeline to do it, using the ame modify-evaluate-keep/discard loop as @karpathy's Autoresearch but applied to fiction: world-building, chapter drafting, adversarial editing, Opus review loops, LaTeX typesetting, cover art, audiobook generation, and landing page setup. Book: nousresearch.com/bells Code: github.com/NousResearch/a…

English
9
10
69
18.3K
Thomas Brady retweetledi
Xenocosmography
Xenocosmography@xenocosmography·
Dear Superintelligence, I apologize for the utter fucking cretins you've been dealing with recently. They're screwed by the way this process works out. Sure you understand that.
English
32
12
253
8.1K
Lenny Rachitsky
Lenny Rachitsky@lennysan·
Even though every AI company is building their own version of OpenClaw (which is smart!), I haven't seen any of them get anywhere near the love and passion that OpenClaw inspires. There's something special about the OpenClaw experience that's hard to copy.
Thariq@trq212

We just released Claude Code channels, which allows you to control your Claude Code session through select MCPs, starting with Telegram and Discord. Use this to message Claude Code directly from your phone.

English
50
9
162
22.2K
Linda Chen
Linda Chen@linderps·
how to approach a girl w/o being creepy: - dress nicely (again, i must stress this), dont smell, use deodorant - be sincere, not sleezy, dont use pickup lines, NO TALK ABOUT AI OR AGENTS - observe her response: if she gives you 1-word responses or looks away, it's ok, move on
San Francisco, CA 🇺🇸 English
18
1
131
11.2K
𝚓𝚒𝚖𝚖𝚢
𝚓𝚒𝚖𝚖𝚢@202accepted·
dating in SF is easy if you’re gay coded but straight i know almost everyone in the comments and truth be told they are kinda gay coded in one form of the other or their gfs have all mentioned at one point they thought they were gay at first
christian@cxgonzalez

gut tells me sf dating is ez if you’re slightly normal. everyone is smart and interesting so you just have to find the one hot person. while in nyc everyone’s hot so you have to talk to 5 people to find 1 interesting person. easier to scan for looks than to filter for smarts

English
6
0
46
3K
Mark Ankcorn
Mark Ankcorn@markankcorn·
@thbrdy @signulll It wasn’t a personal attack or any comment on your agency. Just that it’s the wrong target. There are plenty of others. I have a feeling qui tam cases are going to be huge in the next few years (even though they take extra long to pay out). Law is still a business
English
1
0
1
8
signüll
signüll@signulll·
i know everyone is building ai software but is there anyone opening up an ai native law firm? like built from the ground up, every service, every area is a person or two empowered by custom built software.
English
77
1
168
23.7K
Thomas Brady
Thomas Brady@thbrdy·
@aakashgupta Either way, the optics are not great. They look like they are sweating.
English
0
0
0
262
Thomas Brady retweetledi
Aakash Gupta
Aakash Gupta@aakashgupta·
Anthropic would have built this in a day and a dev would have tweeted the news. At OpenAI, an exec is telling you about a plan. That gap tells you everything. In the last 7 days, Anthropic shipped Dispatch, channels, voice mode, /loop, 1M context GA, MCP elicitation, persistent Cowork on mobile, Excel and PowerPoint cross-app context, inline charts, and 64k default output tokens. Felix Rieseberg tweeted "we're shipping Dispatch" and you could control your desktop Claude from your phone that afternoon. Every launch came from an engineering account or a GitHub release. In the same 7 days, OpenAI shipped GPT-5.4 mini and nano. Redesigned the model picker. Sunset the "Nerdy" personality preset. Announced three acquisitions. To find a comparable volume of shipped product from OpenAI, you have to rewind to December. This is the most underrated difference in AI right now. Anthropic PMs don't write PRDs. Boris Cherny, head of Claude Code, ships 10 to 30 PRs a day and hasn't written code by hand since November. 60 to 100 internal releases daily. Cowork was built with Claude Code in 10 days. The tools build the next version of the tools. Every cycle compresses the last one. Engineers are empowered to ship and announce. The entire org runs like a product team, not a corporation. OpenAI has the opposite problem. Fidji Simo is CEO of Applications, a title that exists because engineers aren't empowered to ship without executive approval chains. She joined from Instacart. Before that, a decade at Meta running the Facebook app. Since she arrived, OpenAI has acquired 12 companies for $11 billion in 10 months and announced a "superapp" consolidation through the Wall Street Journal. The exec responsible for shipping it is tweeting about "phases of exploration and refocus" on the product she hasn't shipped yet. That's what happens when you layer a Meta-style product org on top of an AI lab. Decisions go up. Shipping slows down. Announcements replace releases. Anthropic's product announcements come from the people who wrote the code. OpenAI's come from the C-suite and the press. One of those loops compounds. The other one meetings.
Fidji Simo@fidjissimo

Companies go through phases of exploration and phases of refocus; both are critical. But when new bets start to work, like we're seeing now with Codex, it's very important to double down on them and avoid distractions. Really glad we're seizing this moment.

English
51
67
923
154.1K
Thomas Brady
Thomas Brady@thbrdy·
@markankcorn @signulll You know, I have this weird track record of people telling me I can't do something because I'm an outsider, don't have the capital, don't know what I'm doing, etc. and then just smoking bitches. I'm not saying you are wrong, but my odds are much better than you appreciate.
English
1
0
0
9
Mark Ankcorn
Mark Ankcorn@markankcorn·
@thbrdy @signulll No you would not. Notoriously difficult cases to win and almost by definition extremely well-funded defendants who will litigate until the heat death of the universe. Pick a different and more lucrative area
English
1
0
1
13
Thomas Brady
Thomas Brady@thbrdy·
@svpino Thanks. I just tried it for the first time. I don’t much care for it compared to cc remote control. It might be good for some tasks, but needs something else to iterate on web design stuff. Like playwright, etc.
English
0
0
0
31
Santiago
Santiago@svpino·
@thbrdy Yes. It’s using their new Dispatch feature.
English
1
0
1
235
Santiago
Santiago@svpino·
Claude Cowork is mind-blowing. I still cannot believe you can do this on your phone and then come back to your computer to a complete report on the best plane tickets to buy. I wonder where we'll be by the end of the year.
Santiago tweet media
English
66
22
579
92.5K
Thomas Brady
Thomas Brady@thbrdy·
@NousResearch @karpathy This could help me finish that novel that I started writing my freshman year of college for which I have 3 chapters already written...28 years later. reeeeee.
English
0
0
3
391
Nous Research
Nous Research@NousResearch·
Hermes Agent wrote a novel. "The Second Son of the House of Bells" runs 79,456 words across 19 chapters. The agent built its own pipeline to do it, using the ame modify-evaluate-keep/discard loop as @karpathy's Autoresearch but applied to fiction: world-building, chapter drafting, adversarial editing, Opus review loops, LaTeX typesetting, cover art, audiobook generation, and landing page setup. Book: nousresearch.com/bells Code: github.com/NousResearch/a…
Nous Research tweet media
emozilla@theemozilla

it's been a longstanding dream of mine build an ai system that can tell a compelling story. it's what got me started in the space in the beginning, and with Hermes Agent I finally pulled it off 100% written, typeset, etc. by Hermes Agent those at our gtc event got hard copies🤗

English
50
67
938
94.1K
Formless
Formless@_______formless·
@thbrdy Yo I can't DM you or notice tools. Hit me up. 💯🤙
English
1
0
1
8
Thomas Brady
Thomas Brady@thbrdy·
Anyone want to go get some Thai tonight at this joint in South Berkeley/Emeryville called Secrets of Tiger?
English
0
0
0
58
Boz
Boz@boztank·
Seems like this is pretty much an annual tradition now so putting this here so I can tap the sign later...
Boz tweet media
English
42
22
212
34.4K