Finally got blackout curtains again - and automated opening them every morning with a @SwitchBot curtain bot. Feels like one of the easiest way to increase #sleepquality
Bus companies might be borrowing from airlines’ playbooks - it’s cheaper to buy a @PeterPanBus ticket to BOS/Logan with a stop in Boston than to go to Boston directly
yes things are changing fast, but also I see companies (even faang) way behind the frontier for no reason.
you are guaranteed to lose if you fall behind.
the no unforced-errors ai leader playbook:
For your team:
- use coding agents. give all engineers their pick of harnesses, models, background agents: Claude code, Cursor, Devin, with closed/open models. Hearing Meta engineers are forced to use Llama 4. Opus 4.5 is the baseline now.
- give your agents tools to ALL dev tooling: Linear, GitHub, Datadog, Sentry, any Internal tooling. If agents are being held back because of lack of context that’s your fault.
- invest in your codebase specific agent docs. stop saying “doesn’t do X well”. If that’s an issue, try better prompting, https://t.co/SOjpn47yxo, linting, and code rules. Tell it how you want things. Every manual edit you make is an opportunity for https://t.co/S1ZvtYQwta improvement
- invest in robust background agent infra - get a full development stack working on VM/sandboxes. yes it’s hard to set up but it will be worth it, your engineers can run multiple in parallel. Code review will be the bottleneck soon.
- figure out security issues. stop being risk averse and do what is needed to unblock access to tools.
in your product:
- always use the latest generation models in your features (move things off of last gen models asap, unless robust evals indicate otherwise). Requires changes every 1-2 weeks - eg: GitHub copilot mobile still offers code review with gpt 4.1 and Sonnet 3.5 @jaredpalmer. You are leaving money on the table by being on Sonnet 4, or gpt 4o
- Use embedding semantic search instead of fuzzy search. Any general embedding model will do better than Levenshtein / fuzzy heuristics.
- leave no form unfilled. use structured outputs and whatever context you have on the user to do a best-effort pre-fill
- allow unstructured inputs on all product surfaces - must accept freeform text and documents. Forms are dead.
- custom finetuning is dead. Stop wasting time on it. Frontier is moving too fast to invest 8 weeks into finetuning. Costs are dropping too quickly for price to matter. Better prompting will take you very far and this will only become more true as instruction following improves
- build evals to make quick model-upgrade decisions. they don’t need to be perfect but at least need to allow you to compare models relative to each other. most decisions become clear on a Pareto cost vs benchmark perf plot
- encourage all engineers to build with ai: build primitives to call models from all code bases / models: structured output, semantic similarity endpoints, sandbox code execution. etc
What else am I missing?
Excited to announce @coda_hq’s next chapter: we've entered into an acquisition agreement with @Grammarly!
With @shishirmehrotra taking over as CEO of the newly combined company, we’re looking forward to building the AI-native suite of the future.
https://t.co/gJT1pwFt59
Copilot is likely writing more code on GitHub than humans.
As AI generates more of the world’s code, we need a new developer toolchain to review, test, and deploy quickly & safely. Here’s what we believe it looks like @withgraphite: 🧵
Instant answers, from 500+ tools.
Today, we’re partnering with @SnowflakeDB to launch Coda Brain, your team’s favorite know-it-all. It’s a turnkey AI platform that understands your company’s knowledge & data by syncing your tools into a comprehensive, up-to-date brain...
🧵 1/X
🚨 BREAKING NEWS!
As Snowflake CEO @RamaswmySridhar just mentioned in his Data Cloud Summit opening keynote, Coda and @SnowflakeDB have co-developed an all-new product: Coda Brain!
Coda Brain is a turnkey AI platform that lets you spend less time hunting for information—and more time putting data to work.
Learn more on our blog: https://t.co/sUoPoF33cr
.@USPS’s @USPSWebTools address API appears to be down. No way to reach them because the contact form also uses the API and doesn’t let you contact them without validating your address 🤦♂️
🚨 Huge news!
Today we’re launching a new partnership with @SnowflakeDB to help you turn data into action across the enterprise, and announcing an investment in Coda, led by Snowflake Ventures!
The divide between data and action is one of the key problems facing modern teams.
With Coda and Snowflake, we’re aiming to democratize your team’s data and bring Snowflake’s Data Cloud into the platform where your team gets things done: Coda.
It all starts with the Snowflake Pack, which brings live data into your team’s Coda workspace.
But that’s just the beginning.
Read @shishirmehrotra’s blog post for more on the future of our partnership and Enterprise AI.
https://t.co/dYUpFx9QlY
Vote! The next 24 hours is going to determine the future of San Francisco.
Vote the doom loopers out and the boom loopers in!
Vote for the @SFDemsForChange slate.
🚀 Big news: today we’re excited to launch Coda 4.0 – the all-in-one platform for businesses, powered by AI.
You might be saying to yourself “Coda releases new features every day, what’s so special about Coda 4.0?”
We’re glad you asked...
Hello, Dreamforce! 👋
Join us for a happy hour with Coda, @Workato, @glean, and @10Kview on September 12th at Press Club in San Francisco from 5-9pm for networking and drinks!
RSVP here: https://t.co/rsfgbY5DNC
11 years ago, we launched @kp_fellows to find next-gen leaders. Today, our community is 800+ strong, including founders, execs, and @kleinerperkins investors. Thrilled to expand with KP Next, designed for experienced engineers, PMs, and designers who want to build the future🚀
I asked HustleGPT if it had a message for everyone on Twitter and not gonna lie the grift is real.
Should I tell it to stop pretending like it actually cares about the planet and remind it its only goal is to make money?