Robert

[Robert @Tantalon]

Love how Pointer AI handles Google Docs suggestions - way more intuitive than Gemini. Just wish it supported tabs and external context. https://t.co/0AZ9kGCVL8 @whoisrade

[Jafar Najafov @JafarNajafov]

Wild: Anthropic just proved that prompt engineering is dead. Context engineering is the new game, and it changes everything about building AI agents 🤖

For years, we obsessed over the perfect prompt.

The right words. The perfect phrasing. The magic sentence that makes GPT-4 finally understand what you want.

But Anthropic just dropped a reality check:

Prompts are the easy part. Context is the bottleneck.

Here's what they found:

→ LLMs have an "attention budget" like human working memory, but for tokens
→ More context = worse performance (they call it "context rot")
→ The transformer architecture creates n² relationships between tokens
→ Every token you add depletes focus exponentially

Translation? Stuffing your AI with information is like trying to focus while 50 people scream at you.

So what's the new playbook?

↳ Just-in-time retrieval
- Instead of loading everything upfront, agents pull data dynamically using tools
- Claude Code does this brilliantly: it uses grep, head, tail to analyze databases without ever loading full files into context

↳ Compaction
- Summarize the conversation when you hit limits
- Keep architectural decisions, discard redundant tool outputs
- Maintain continuity without context bloat

↳ Structured note-taking
- Agents write persistent notes outside the context window
- Pull them back only when needed
- Think: AI maintaining its own https://t.co/xZDZxFLBJK file

↳ Sub-agent architectures
- Specialized agents handle focused tasks
- Return compressed summaries (2k tokens) instead of raw exploration (50k+ tokens)
- Separation of concerns at the AI level

The shift is profound:

Before: "How do I write the perfect prompt?"
Now: "What's the minimal high-signal context that drives my desired outcome?"

This isn't just academic.

Goldman Sachs is using AI to draft S1s.
Perplexity is automating financial analysis.
Claude Code is refactoring entire codebases.

None of this works if you drown the AI in context.

The best AI engineers aren't prompt wizards anymore.

They're context architects.

[Jozsef Czapovics @jczapovics]

I looked at our dev team's last 30 days of @windsurf_ai stats. I expect to get that code written percentage up to 90 %+ this year.

Total lines of code written by Cascade: 631,591
% of new code written by Tab: 47%

Cascade messages sent: 20,957 (Write: 17,534, Chat: 3,414)

Tool calls: 89,420 (Code edits: 29,300)

Model usage for reasoning in Cascade:
Claude 3.7 Sonnet: 64%
Claude 3.5 Sonnet: 15%
GPT-4.1 (promo): 12%

Acceptances by Tab: 38,436

[Robert @Tantalon]

@im_roy_lee @trycluely Great decision. We had this challenge 8 months ago. Everyone recommended a loopback. UX was terrible. Moved to ScreenCaptureKit - it fixed all our problems.

[Paul Couvert @itsPaulAi]

So Google Sheets now has a "=AI" formula?!

You can process data that was impossible before in a spreadsheet.

Gemini understands what's in the cells and returns a tailor-made answer according to your instructions.

Examples and formulas below

[Wise @trikcode]

Your job isn’t to write code…it’s to solve problems.

[Krystian Lieber @KrystianLieber]

Curious why the @windsurf_ai Team Plan lacks the automatic command execution feature available to individual users. Is this feature coming to teams soon? #WindsurfFeatures #TeamPlan @codeiumdev

[Emiliano @Emiliano_GLopez]

As a programmer, you have to make sure not to become too reliant on AI. Sometimes Claude Code won't work, so you have to know how to use Cursor.

[Matthew Berman @MatthewBerman]

#1 Trending Github Repo: Screenshot-to-Code

A simple tool to convert screenshots, mockups and Figma designs into clean, functional code using AI

Check out the WILD demos below 👇

[Robert @Tantalon]

Is @vestapp_ai coming back? It was way better than me at picking stocks!

[Ricardo mahfoud @RicardoMahfoud]

Need a quick demo of the best Cursor features?

Watch my workflow where I create a native macOS app using Cursor Composer in under 8 minutes without writing a single line of code.

[Kyle Corbitt @corbtt]

Wait did OpenAI just drop the price of GPT-4o (the real one, not mini) by 50% without even announcing it?

Like there is nothing on their blog/@OpenAIDevs , just a new pricing page with $2.50/1M tokens listed. Bullish for OpenAI. 🤯

[Pietro Schirano @skirano]

Introducing Claude Engineer 2.0, with agents! 🚀

Biggest update yet with the addition of a code editor and code execution agents, and dynamic editing.

When editing files (especially large ones), Engineer will direct a coding agent, and the agent will provide changes in batches.

Batches are smartly selected based on file complexity.

The code execution agent will run the code and check for issues.

It can even start processes (like live servers) and end them.

It's insanely powerful! 🔥

[Robert @Tantalon]

@deedydas We now do a first draft using LATS, followed by ReAct agents to refine the output. From this, we generate:
- A product sheet for humans, with an overview, modules, and instructions
- A list of technical decisions, used as context in AI tools to tailor the code to the product

[Robert @Tantalon]

@deedydas Thanks for sharing your insights on understanding codebases! We acquire software companies, so we constantly face this. We started the same way — read all code into context; it’s simple and quick. However, we soon found better results with an agentic approach.

[Ethan Mollick @emollick]

This remains one of the most consequential experiments in AI: Bloomberg trained a GPT-3.5 class AI on their own financial data last year…

…only to find that GPT-4 8k, without specialized finance training, beat it on almost all finance tasks.

Hard to beat the frontier models.

[Robert @Tantalon]

Powerful AI assistant you can take with you on the go, I can't wait to see what we can do with this!

[Interpreter @OpenInterpreter]

Introducing the 01 Developer Preview.

Order or build your own today: https://t.co/Z6puIyGwqo

The 01 Light is a portable voice interface that controls your home computer. It can see your screen, use your apps, and learn new skills.

This is only the beginning for 01— the open-source foundation for this new era of AI devices.

[Burny - Effective Curiosity @burny_tech]

Researchers gave AI an 'inner monologue' and it massively improved its performance | Scientists trained an AI system to think before speaking with a technique called QuietSTaR. The inner monologue improved common sense reasoning and doubled math performance
Quiet-STaR: Language Models Can Teach Themselves to Think Before Speaking

[Jeff Dean @JeffDean]

We're starting to roll out API support for Gemini 1.5 Pro for developers. We're excited to see what you build with the 1M token context window!

We'll be onboarding people to the API slowly at first, and then we'll ramp it up. In the meantime, developers can try out Gemini 1.5 Pro in the AI Studio UI right now:

https://t.co/HG3inVgrc7

[Pietro Schirano @skirano]

Claude Opus's ability to orchestrate subagents is absolutely insane and deserves more attention.

Watch Claude direct subagents to build an entire drawing app. 🧙‍♂️

If you like, I could share this code, which can basically solve any goals you present, step by step.

Like Devin!