Senthil

Be kind. Live with integrity and take nothing personally.

[Min Choi @minchoi]

This is actually useful.

LangChain just released OpenWiki.

It's an open-source agent that creates a wiki for your codebase, connects it to your coding agent, and keeps it updated as your repo changes.

Your AI coding agent gets long-term repo context without stuffing everything into CLAUDE.md.

Here's how to setup. Save this.

[Gergely Orosz @GergelyOrosz]

I’m starting to realize just how important it is to understand context sizes, context rot, context compression & similar behaviors to understand why these models often fall short

Eg why it is that you give it a large block of stuff and the model “forgets” about parts of it etc

[The Pragmatic Engineer @Pragmatic_Eng]

Anyone who thinks software engineering is ‘going away’ doesn’t understand the job. @KentBeck, creator of XP and TDD, on why Dario has it wrong:

[Gergely: Dario said, I quote, ‘coding is going away first, then all of software engineering’. ]

“That's a statement by someone who doesn't understand software engineering. Coding is part of what you're doing, but it's only a small part of what you're doing, even if it takes up a fair amount of time.

You're building confidence, you're building connections with other people, you’re building your own understanding. All those things are happening while you're coding. And coding's actually a great way to cement understanding. The more you program, the more you understand the domain that you're working in. And so to say, well, we're just going to pass all that off to a machine. Well, that's not all there is to it.

A couple of days ago I saw a phrase, and it really hit me, that we're accumulating code faster than we're accumulating trust now. And that sense of trust comes from me struggling to understand some domain concept, ah, I get it! I represented it in the code. I write tests that demonstrate that I really did understand it and now, I trust my program. If we're programming together, that act of programming together means that we trust each other more.

And none of that can be automated. None of that occurs. If we prompt, we get the finger guns, the genie goes, yeah, it's all finished, boss. And it is like, well, hang on, finished. What's finished?”

[Satya Nadella @satyanadella]

The future of the firm is a learning loop in which human capital and token capital compound.

With our new Frontier Co., our ambition is to help every enterprise build its own AI capability, and to help create a frontier ecosystem where every organization can turn its knowledge, workflows, and judgment into its own AI systems that continuously improve. https://t.co/mvYhkRFyqa

[Werner Vogels @Werner]

I am not a big fan of doing this client-side, for a whole slew of reasons, but this is impressive engineering by the @Zalando team. I like the "lessons learned" they conclude with. https://t.co/5kZkLsT2ro

[Senthil @senthil_l]

The narrative fallacy describes our tendency to stitch unrelated facts and random events into neat, oversimplified cause-and-effect stories. Humans do this a lot, so are the GenAI models that are basically predicted the next token from human generated content.

[Reads with Ravi @readswithravi]

In these uncertain times, Becoming Yourself by Suzuki Roshi is an extraordinary gift and the perfect book for sanity, balance, and guidance. Through simple yet profound wisdom, it encourages a life of mindfulness, compassion, and non-attachment.

5 lessons from the book:

[Simon Willison @simonw]

Notes (and a Pelican) on Claude Sonnet 5 - the new tokenizer makes it ~1.4x more expensive for English, ~1.33x more expensive for Spanish but roughly the same price for Simplified Mandarin https://t.co/UUnmjtPaSi

[Gergely Orosz @GergelyOrosz]

This could well end quicker than most people assume, because coding agents in the cloud are coming, and they are coming fast, especially inside places like Cursor (I was in their offices yesterday, and local agents will prob go away soon as I read the room)

[Andrew Ng @AndrewYNg]

“Loop engineering” is a hot buzzphrase after mentions of it by Boris Cherny (Claude Code’s creator) and Peter Steinberger (OpenClaw's creator) went viral on social media. Loops are now a key part of how we get AI agents to iterate at length to build software. In this letter, I’d like to share my 3 key loops, shown in the image below, for building 0-to-1 products. These loops guide not just how I build software, but also how I decide what software to build.

Agentic coding loop: Given a product specification and optionally a set of evals (that is, a dataset against which to measure performance), we can have an AI agent write code, test its work, and keep iterating until the code is bug-free and meets its specification. This idea of closing the loop took off around the end of last year, and it has been a game changer in enabling coding agents to work longer productively without human intervention. For example, over the weekend, I was building an app for my daughter to practice typing, and my coding agent could easily work for around an hour, using a web browser to check what it had built multiple times before getting back to me, without needing my intervention.

The engineering loop executes quickly. Every few minutes, the coding agent might build and test a new version of the software. I hear frequently from developers who are finding new ways to engineer more effective engineering loops. This is an active area of invention!

Developer feedback loop: In this loop, a developer examines the current product and steers the coding agent to improve it. Last year, a lot of developers (including me) were acting as the QA (quality assurance) function for our coding agents, manually finding bugs and then asking the agent to fix them. But with coding agents much more able to test their own code, the amount of time we need to spend on this function has decreased significantly. This allows us to make higher-level product decisions, such as what key features to offer, where the UI needs improvement, and so on.

The developer-feedback loop operates over time intervals between tens of minutes and hours — that's how frequently a developer might review a product and give feedback. In the case of the typing app, I changed my mind a few times about the visual design, what cat costumes she can unlock as she learns (she loves cats), and the user flow for a grown-up to log in and steer the child's learning experience.

When a developer has a clear vision for what to build, it is still a lot of work to translate that vision into a specification for a coding agent to implement. Further, after the developer has seen an implementation, they might update (or perhaps clarify) the spec to steer it toward what they want. If you find that the system repeatedly runs into certain problems, building a set of evals for the agent becomes useful.

AI-native teams are increasingly using AI to help shape product direction, for example, automating the gathering and analysis of usage data, summarizing written and verbal customer feedback, or carrying out competitive analysis. However, for pretty much all the products I’m involved in, I see humans as having a significant context advantage over current AI systems — we know a lot more than the AI system about the users and the context the product has to operate in — and thus humans play a critical role. Many people describe this human contribution as “taste,” but I prefer to think of it as humans having a context advantage, since that gives us a clearer path to helping AI systems get better. This also speaks to why this step can’t be automated: So long as the human knows something the AI does not, human-in-the-loop is needed to to inject that knowledge into the system.

External feedback loop: This includes a wide range of tactics like asking a few friends for feedback, launching to alpha testers, or putting the code into production with A/B testing. These tactics are usually slow, rarely taking less than hours and sometimes taking days or even weeks. This data informs the developer vision, which in turn continues to drive the detailed product spec, which in turn drives the coding agent.

With coding agents speeding up software development, more engineers are starting to play a partial product management role. For many engineers who are growing into this role, the hardest part is shaping the product vision and striking a balance between building (bridging the gap between vision and spec) and getting user feedback to evolve the vision. It is important to do both!

I will write more about how to do this in future posts, but for now, I find it encouraging that engineers are playing an expanded role (just as product managers and designers now do more engineering).

[Original text: The Batch]

[Werner Vogels @Werner]

Two-pizza culture was never about feeding hungry engineers. It was about ownership, speed, and not being bogged down by bureaucracy. Working backwards from the customer and putting pen to paper have always been our way of forcing clarity of thought. But the industry is changing, and it’s time to reconsider how we bring products to live. Read more: https://t.co/TIfpLLWRNQ

[Karthik Hariharan @hkarthik]

Manually compacting your LLM sessions is annoying. Feels like manually freeing up memory, manually rebuilding database indexes, or manually freeing up disk space.

True AGI that really fulfills the promise of AI won’t require this.

[Sridhar Vembu @svembu]

American Treasury Secretary Scott Bessent, in a major speech, spells out what we in India would call their own "Swadeshi economic policy".

The first principle he states: "The first is that economic security begins with national capacity… The nation that depends on its adversaries for critical inputs is not truly sovereign. And the nation that reduces its economics to consumption is not truly prosperous."

Powerful words. I agree with him and India must resolutely apply the same principles. East Asian nations are all strongly Swadeshi too.

I am proud to be closely associated with the Swadeshi Jagran Manch, and SJM has long advocated similar principles, for over 3 decades now.

[Scott Hanselman 🌮 @shanselman]

"container images" for windows...out of the box, no extra software needed.

[Steve Yegge @Steve_Yegge]

So we looked at the GitHub numbers, and Beads has over 650k downloads. Beads is still absolutely the best way to work with coding agents, all flavors. It works with your workflow, and there's almost nothing to learn. It just works. I'll be posting a series of tips on how to use Beads to do @bcherny's loops, both small and large.

https://t.co/PQU1TNJWfQ

[Gergely Orosz @GergelyOrosz]

What happens when the most capable coding model (Fable / GPT-5.6) is banned by the US government, and the sending most capable is an open model?

What happens: a LOT of businesses and devs simply move to use the open model (GLM-5.2) via inference providers. Cheaper + better!

[Deedy @deedydas]

WhatsApp may be the statistically best app we've ever made.

It's a product manager's dream with its unparalleled addictiveness (DAU / MAU) of 87% and stickiness (M1 retention) of 86%, both #1 in the world while having the #1 most monthly active users for any non preinstalled app of 2.7B users.

Here are the top 25 most used apps in the world by MAU on both these metrics. Some surprising observations:
— There are now 15 1B+ user apps in the world, 8 by Google, 4 by Meta.
— The 3 that aren't are TikTok, Telegram (!) and ChatGPT
— Telegram has more users than Spotify, Pinterest, Netflix, Amazon, Snap!
— ChatGPT's one month retention is #5 after WA, Instagram, Chrome and Youtube. 2yrs ago, it was been #20 by M1 retention
— Shopee, a shopping app in southeast Asia, is huge and retains users better than Amazon!

Useful way to break down consumer businesses especially within certain categories (X vs Reddit vs Threads is a good one). It's shocking how few new apps have been able to break through in the past 10yrs.

[Steve Yegge @Steve_Yegge]

Ubering into downtown San Francisco, every ad looks like desperation. All agent stuff nobody will use. Haven't seen anything that a CIO should or would buy. It's all starting to look like the same ad.

[Sridhar Vembu @svembu]

Databricks is embracing Chinese open source model GLM-5.2.

Earlier, Microsoft said they were looking to embrace DeepSeek.

Major American software companies are embracing Chinese open source models to cut token costs.

I have long advocated using the Chinese models running in one's own infrastructure.

We have Zoho-specific small models that we use in production. This approach also will become common for larger companies, as it becomes easier to set up the training pipeline to train your own org specific model.

[Boris Cherny @bcherny]

As engineering, product, design, DS, etc. melt into a new kind of role, I was reflecting on what roles might look like in the future. For example, when I look at the Claude Code team I see what I think is five archetypes:

1. Prototyper: comes up with brand new ideas; churns out many ideas, most of which don't ship
2. Builder: quickly turns a prototype/idea into production-grade product/infra
3. Sweeper: cleans up the UI, simplifies the code and system, unships, optimizes performance
4. Grower: takes a product that has been built and iterates on it to improve Product-Market Fit
5. Maintainer: owns a mature system to make it secure, reliable, fast, and efficient as it scales

Many people span across 2 roles, and sometimes 3 roles. I also notice that these roles are not really tied to job function -- eg. across Anthropic, some designers match category 1, some 2, some 3; same for engineers, PM, DS.

A healthy team needs a mix of these, depending on the product:

- A product that is new and pre-PMF needs people that are strong at 1+2+3
- A product that is growing and has found PMF needs 2+3+4 and some 5
- A product that has strong PMF needs 3+4+5 and some 2

Maybe product roles of the future will look more like this, and less like the domain-specific roles of today?