pdash

@levie That's exactly what we do at exceeds.ai. From prompt to commit to pr to deploy - we measure $ spent by an engineer, and put it on the board with business outcomes

A common trend emerging in larger enterprises is token budgeting as a major topic. As agents can do more and more long running tasks, and thus take vastly more compute, allocation of tokens across teams becomes a very real thing in the enterprise. Companies spend a meaningful amount of time deciding how much to spend on talent, marketing campaigns, events, laptop setups, and even the cost of lunches. Tokens will be no different. Tokens will similarly need to be excruciatingly well-managed because you’ll need to ensure you don’t blow up your budget, and you’ll need to ensure that the tokens are flowing to the highest and most useful parts of work. You don’t want to find out you burned your monthly budget on something relatively low value and then be blocked on the much higher value task later. Doing this at large company scale is extremely hard as you have layers of abstraction on data and visibility into the digital work being done by agents in any central way. This is going to mean that agentic spend will increasingly will expand beyond the confines of the IT budget, and end up in organizational budgets like other expenses. Ultimately team and org leaders will have to be given budgets for this, but even they don’t have adequate visibility and controls in most cases. We’ll need all new software just to solve this problem, and it’s probably an opportunity for startups in its own right. Going to be an all new era of enterprise resource allocation, especially while we compute constrained.

“History doesn’t crawl; it leaps.” - nnt "How did you go bankrupt?" Bill asked. "Two ways," Mike said. "Gradually and then suddenly." - Hemmingway

@MrMikeInvesting Funny how sandisk and micron had/have the same founder/ceo

We are currently in a “once in a lifetime” AI super cycle… Phase 1 was: (already gone) Semiconductors ~ $NVDA, $AMD, $INTC, $ARM Phase 2 is: (passing by now) Memory ~ $MU, $SNDK, $WDC Photonics ~ $AAOI, $AEHR, $LITE, $MRVL The current phase is Neo Cloud/AI infrastructure: $IREN, $NBIS, $CRWV, $CIFR, $APLD Next wave (many will miss) Rare Earths ~ $USAR, $MP, $UUUU, $FCX Power & Cooling~ $VRT, $CEG, $OKLO, $OSS Finally it all concludes with these 3 sectors: Robotics ~ $TSLA, $PATH, $SERV Space ~ $RKLB, $ASTS, $PL, $LUNR Drones ~ $ONDS, $AVAV, $LMT Many will make generational wealth from this AI super cycle over the next 7 months. Save this to look back on later…

@gdb Great in cursor

how is gpt-5.5 performing for you?

@OpenAIDevs Ooh elixir

📣 What if every open issue had a Codex agent? That’s the idea behind Symphony, an open-source agent orchestrator for Codex that turns task trackers into always-on systems for agentic work, letting humans focus on review and direction.

@retailtraderAI Invalidation vs stop loss is quite interesting

@garrytan CXO’s - you can’t manage what you can’t measure. exceeds.ai

If you are in startups or tech It is literally cook or be cooked

@SidChoudhury Headcount budgets -> Token budgets

@pradeepta Perfect bug to have when you charge your customers by tokens 🙂

Oh the amount of Claude tokens we burn by analyzing leaked Claude Code source code. No wonder the Claude team is chill about it 😎

@judthoughts @pmarca I am

@pmarca I am.

“Who is John Galt?”

@karpathy Yes. This has worked great - I always ask it to present counter arguments against its own pov

- Drafted a blog post - Used an LLM to meticulously improve the argument over 4 hours. - Wow, feeling great, it’s so convincing! - Fun idea let’s ask it to argue the opposite. - LLM demolishes the entire argument and convinces me that the opposite is in fact true. - lol The LLMs may elicit an opinion when asked but are extremely competent in arguing almost any direction. This is actually super useful as a tool for forming your own opinions, just make sure to ask different directions and be careful with the sycophancy.

@SVG__Collection One of the best

“Getz/Gilberto” by Stan Getz & João Gilberto was recorded 63 years ago today.

@JeffDean @prajdabre @GhemawatSanjay Thank you for MR. We read the paper and build yahoo execution framework in Perl for yahoo! Ad system (pre hadoop). Lots of fun!

Main lore around the origin of MapReduce was that we were rewriting our indexing pipeline for the search system, and we realized that lots of the different phases were conceptually simple but required large scale processing (extract link text from each page, identify language for each page, compute checksum of contents to identify duplicates, etc). Each phase needed to be parallelized, made robust to machine failures, etc. Squinting at each of the phases we came up with MapReduce as an abstraction where we could have an implementation that would do all the complex work under the abstraction boundary, and where the expression of the operations could be nice and simple.

Btw this question is a homage to @JeffDean and @GhemawatSanjay, who wrote the paper on MapReduce. I wonder if there was some fun lore around this.

@antirez Ah thank you

If you explain how to do it, Claude Code is able to use Codex when it is incapable of solving certain issues. This way you have the best of both worlds. Here is the skill file that lets Claude use Codex. gist.github.com/antirez/2e0772…

@lennysan ChatGPT is unparalleled for 5.2 for getting well architected backends done right

What's the most underrated/underhyped AI tool in your stack right now?

@t_blom txt2post.com/pradeepta

@t_blom txt2post.com

I really want a text editor that is voice-native and makes inline changes like a code editor. So I built one

@mattvanswol I’m sorry Matt. Larry (my rescue chihuahua) send you nuzzles

Struggling to not cry while writing this. Sadly, our 10 year old Golden Retriever, Winston, has been diagnosed with an aggressive form of bone cancer. It is not survivable. I got Winston at the lowest point in my life, and in so many ways, he saved me. I was an alcoholic, single, so so depressed. I was a bitter, angry, selfish person. ...and I knew it So upon recommendation of a friend, I got Winston. And he changed my fucking life. I learned structure. I learned how to talk to strangers. I went outside every day on runs and walks. He gave me the discipline to wake up early. He gave me a reason to not stay out late. I learned that doing the right thing daily is what changes you, slowly. And because Winston needed me, I started becoming someone worth needing. But then, he didn't just become my dog. He became my wife's dog. My kid's dog. And he loved them with the passion of a thousand fiery suns. And I don’t think I’ll ever be able to explain what it means to watch a dog learn to love your kids as much as you do. He's the greeter. He's my shadow. A peacekeeper. The one constant, no matter what kind of day we’ve had. I'm gonna miss you so much.

@flowidealism You can’t bet without finding out what is each NOT telling you

Cal Newport asks: Imagine you're an admissions officer comparing two applicants. David is captain of the track team and took Japanese calligraphy. Steve does marketing for a sustainability NGO and lobbied delegates at the UN climate conference in Johannesburg. Most people choose Steve because most people can't imagine achieving what Steve achieved in high school. Purpose-driven teens who pursue real projects create applications that stand out from the endless parade of 4.0 GPAs, varsity sports, and volunteer hours.