Stefan Azarić

PM @Microsoft, working on Microsoft Fabric - SQL Data Warehouse!

[Cristian Tăbăcitu @tabacitu]

Beautiful & captivating. If you care about your privacy on the internet - you’ll like this. Read it top to bottom.

https://t.co/eBxnr1vRYi

[Stefan Azarić @azaricstefan]

@GergelyOrosz Just got this post as a notification 😂

[Evan Boyle @_Evan_Boyle]

Remote control is now live in the Copilot CLI!

Run Copilot on your local machine, leave it plugged in, head to the beach, and drive the agent from your mobile phone 😎

[Andrej Karpathy @karpathy]

A few random notes from claude coding quite a bit last few weeks.

Coding workflow. Given the latest lift in LLM coding capability, like many others I rapidly went from about 80% manual+autocomplete coding and 20% agents in November to 80% agent coding and 20% edits+touchups in December. i.e. I really am mostly programming in English now, a bit sheepishly telling the LLM what code to write... in words. It hurts the ego a bit but the power to operate over software in large "code actions" is just too net useful, especially once you adapt to it, configure it, learn to use it, and wrap your head around what it can and cannot do. This is easily the biggest change to my basic coding workflow in ~2 decades of programming and it happened over the course of a few weeks. I'd expect something similar to be happening to well into double digit percent of engineers out there, while the awareness of it in the general population feels well into low single digit percent.

IDEs/agent swarms/fallability. Both the "no need for IDE anymore" hype and the "agent swarm" hype is imo too much for right now. The models definitely still make mistakes and if you have any code you actually care about I would watch them like a hawk, in a nice large IDE on the side. The mistakes have changed a lot - they are not simple syntax errors anymore, they are subtle conceptual errors that a slightly sloppy, hasty junior dev might do. The most common category is that the models make wrong assumptions on your behalf and just run along with them without checking. They also don't manage their confusion, they don't seek clarifications, they don't surface inconsistencies, they don't present tradeoffs, they don't push back when they should, and they are still a little too sycophantic. Things get better in plan mode, but there is some need for a lightweight inline plan mode. They also really like to overcomplicate code and APIs, they bloat abstractions, they don't clean up dead code after themselves, etc. They will implement an inefficient, bloated, brittle construction over 1000 lines of code and it's up to you to be like "umm couldn't you just do this instead?" and they will be like "of course!" and immediately cut it down to 100 lines. They still sometimes change/remove comments and code they don't like or don't sufficiently understand as side effects, even if it is orthogonal to the task at hand. All of this happens despite a few simple attempts to fix it via instructions in CLAUDE . md. Despite all these issues, it is still a net huge improvement and it's very difficult to imagine going back to manual coding. TLDR everyone has their developing flow, my current is a small few CC sessions on the left in ghostty windows/tabs and an IDE on the right for viewing the code + manual edits.

Tenacity. It's so interesting to watch an agent relentlessly work at something. They never get tired, they never get demoralized, they just keep going and trying things where a person would have given up long ago to fight another day. It's a "feel the AGI" moment to watch it struggle with something for a long time just to come out victorious 30 minutes later. You realize that stamina is a core bottleneck to work and that with LLMs in hand it has been dramatically increased.

Speedups. It's not clear how to measure the "speedup" of LLM assistance. Certainly I feel net way faster at what I was going to do, but the main effect is that I do a lot more than I was going to do because 1) I can code up all kinds of things that just wouldn't have been worth coding before and 2) I can approach code that I couldn't work on before because of knowledge/skill issue. So certainly it's speedup, but it's possibly a lot more an expansion.

Leverage. LLMs are exceptionally good at looping until they meet specific goals and this is where most of the "feel the AGI" magic is to be found. Don't tell it what to do, give it success criteria and watch it go. Get it to write tests first and then pass them. Put it in the loop with a browser MCP. Write the naive algorithm that is very likely correct first, then ask it to optimize it while preserving correctness. Change your approach from imperative to declarative to get the agents looping longer and gain leverage.

Fun. I didn't anticipate that with agents programming feels *more* fun because a lot of the fill in the blanks drudgery is removed and what remains is the creative part. I also feel less blocked/stuck (which is not fun) and I experience a lot more courage because there's almost always a way to work hand in hand with it to make some positive progress. I have seen the opposite sentiment from other people too; LLM coding will split up engineers based on those who primarily liked coding and those who primarily liked building.

Atrophy. I've already noticed that I am slowly starting to atrophy my ability to write code manually. Generation (writing code) and discrimination (reading code) are different capabilities in the brain. Largely due to all the little mostly syntactic details involved in programming, you can review code just fine even if you struggle to write it.

Slopacolypse. I am bracing for 2026 as the year of the slopacolypse across all of github, substack, arxiv, X/instagram, and generally all digital media. We're also going to see a lot more AI hype productivity theater (is that even possible?), on the side of actual, real improvements.

Questions. A few of the questions on my mind:
- What happens to the "10X engineer" - the ratio of productivity between the mean and the max engineer? It's quite possible that this grows *a lot*.
- Armed with LLMs, do generalists increasingly outperform specialists? LLMs are a lot better at fill in the blanks (the micro) than grand strategy (the macro).
- What does LLM coding feel like in the future? Is it like playing StarCraft? Playing Factorio? Playing music?
- How much of society is bottlenecked by digital knowledge work?

TLDR Where does this leave us? LLM agent capabilities (Claude & Codex especially) have crossed some kind of threshold of coherence around December 2025 and caused a phase shift in software engineering and closely related. The intelligence part suddenly feels quite a bit ahead of all the rest of it - integrations (tools, knowledge), the necessity for new organizational workflows, processes, diffusion more generally. 2026 is going to be a high energy year as the industry metabolizes the new capability.

[Reid Hoffman @reidhoffman]

.@Replit CEO @amasad on why product managers are the best vibe coders:

[Stefan Azarić @azaricstefan]

Opus 4.5 is a real game changer.

Paired with Copilot in @code it produces high quality code.

[Pierce Boggan @pierceboggan]

@burkeholland @code Agree! My daily driver now

There's a ton of work that goes on beyond the scenes with offline evaluations and our model partner friends to make models great in Copilot.

We should do a blog post on it :)

[Stefan Azarić @azaricstefan]

Absolutely, I agree 100% with Gergely.

[Gergely Orosz @GergelyOrosz]

I keep being amused by devs who assume AI is worthless and useless and is not used to produce production software.

In the comments there are already so many examples. And if you need ones with ~90%+ AI generation that’s used by tens of millions, generating $1B+: Claude Code

[Thomas Dohmke @ashtom]

The evidence is clear: Either you embrace AI, or get out of this career.

Our latest field study with 22 developers who are integrating AI deeply into their workflows reveals a striking trend: those who persist beyond early skepticism emerge with dramatically higher ambition, technical fluency, and job satisfaction. They’re not writing less code – they’re enabling more complex, system-level work through orchestration. And that’s as true for educators as it is for developers themselves.

This shift isn’t hypothetical. It’s happening now. Developers move through clear adoption phases – from dabbling skeptics to strategic AI collaborators – and those who reach the final stage say their identity as developers has transformed. Their focus is no longer on producing code, but on designing systems, directing agents, and validating outputs. “My next title might be Creative Director of Code,” one developer told us. That’s not hyperbole – it’s real. Here’s what we’re seeing:

• AI is on track to write 90% of code within the next 2–5 years.

• Developers aren’t worried. They’re optimistic and realistic about the changes ahead.

• New skills matter now, e.g. agent orchestration, iterative collaboration, and critical verification.

• Time savings? Sure. But the real shift is ambition. Developers are raising the ceiling, not just lowering the cost.

This also has massive implications for education. Teaching syntax alone is obsolete. Students must now learn to guide AI, critique its work, and think across disciplines. Assessments should measure collaboration with AI, not isolation from it.

This is no longer a question of productivity. It’s a question of reinvention. The job of the software developer isn’t disappearing. It’s being reborn. ✨ https://t.co/7oGprPAUb1

[Alex Vacca @itsalexvacca]

Before AWS existed, one company ran the servers for Twitter, LinkedIn, and Facebook's entire app ecosystem.

They owned Node.js, invented containers 8 years before Docker, and Peter Thiel even backed them.

Then something happened...

[Pierce Boggan @pierceboggan]

GitHub Copilot has crossed 20 million users!

it's been a huge year as we evolved VS Code to be the open source AI editor:

- agents: local and remote
- tools: built-in tools for all parts of the editor, like Jupyter notebooks
-mcp: the only editor to support the full MCP spec
- customizable: works like you do with instructions, prompts, and modes
- choice: anthropic, openai, and gemini with bring your own key for even more
- code authoring: completes your current thought and predicts your next

lots more to come! keep the feedback rolling, and building with us in the open :)

[Naval @naval]

If AI lets non-developers replace junior developers, imagine what it lets junior developers do.

[GitHub @github]

Need help with tech debt, bug fixes, and new features? Put GitHub Copilot coding agent to work in the background while you focus on something else, now with more predictable and more affordable pricing. 🚀

Check out our latest update.👇
https://t.co/MLczPPaxEM

[Stefan Azarić @azaricstefan]

@code ❤️ #VS50M #sweepstakes

[Naval @naval]

Only people who have taken products from zero to one are qualified to run a tech company.

[Juan M. Lavista Ferres @BDataScientist]

AI advancements have led us to a breakthrough moment for finding cancer cures. @Microsoft is proud to support The Cancer AI Alliance to help @fredhutch @DanaFarber @MSKCancerCenter @hopkinskimmel accelerate their lifesaving work with the help of AI.

https://t.co/sSjlyohLSG

[Filip Popović @FilipPop_MSFT]

Long awaited!
Secure access to your sensitive data in Microsoft Fabric by providing network isolation and applying required controls on your inbound network traffic!
https://t.co/ILBRW32WpX
#MicrosoftFabric #Microsoft #analytics #warehouse

[Slobodan Marković @smarkovic]

Dame i gospodo – real-time GTFS za Beograd! :)

[Stefan Azarić @azaricstefan]

Do you remember when you joined X? I do! #MyXAnniversary