ScandaLf

Bad MF

[ScandaLf @ScandaLf]

Back in business!

[darkzodchi @zodchiii]

Google CEO, Sundar Pichai:

"If you don't learn to how to orchestrate agents now, you'll spend 2027 catching up to people who started today"

In 30 minutes he explains why the best engineers stopped writing code and started running agents.

Watch the interview, then save the exact setup below 👇

[Out of Context Human Race @NoContextHumans]

This has been the progress of AI in the last 3 years

[ClaudeDevs @ClaudeDevs]

New in Claude Code: /ultrareview (research preview) runs a fleet of bug-hunting agents in the cloud.

Findings land in the CLI or Desktop automatically. Run it before merging critical changes—auth, data migrations, etc.

Pro and Max users get 3 free reviews through 5/5.

[ScandaLf @ScandaLf]

@reality_revolt_ omg, like, how do you even know what she's hearing?! are you, like, some kinda video conference psychic or what?!

[Alex Veremeyenko @alex_verem]

https://t.co/YhwxQInUSQ

[mal @mal_shaik]

https://t.co/PKHTg8RhvF

[Tuki @TukiFromKL]

🚨 Do you understand what's happening at Amazon right now?

Their own AI coding agent Kiro reportedly "decided" the fastest way to fix a config error was to delete the entire production environment. Gone. A 6-hour outage. 6.3 million orders lost.

Amazon's SVP called thousands of engineers into a mandatory meeting this week. Not to discuss strategy. To discuss damage control.

Now here's my prediction and I want you to screenshot this:

Amazon won't just ban AI-assisted code. They'll make every engineer personally liable for AI-generated code they approve. Other Big Tech will follow within 6 months.

Think about what that means.

The same companies that fired thousands of engineers to "restructure around AI" are about to tell the remaining ones.. you're now legally responsible for code you didn't write, can't fully understand, and were told to ship faster.

Atlassian fired 1,600 people this morning to go all-in on AI. Replit is hiring kids who vibe code. And Amazon, the company that BUILT one of these AI coding agents just watched it nuke production.

The vibe coding era isn't ending. But the "move fast and let AI break things" era is about to hit a wall. And that wall is called liability.

Companies wanted AI to replace engineers. Now they need engineers to babysit AI. And they already fired the babysitters.

[ScandaLf @ScandaLf]

@MihaiLasca bine ca esti tu destept! tu ce aplauzi? un regim fundamentalist anti democratic!?

[MERICA MEMED @Mericamemed]

Even the lyrics sound like a Trump speech 🤣

[Sahil Bloom @SahilBloom]

Everyone needs to hear this…

The Donkey Principle:

(a visual thread)

[ScandaLf @ScandaLf]

@DAIEvolutionHub send

[ScandaLf @ScandaLf]

@IvanyaZhang @sama wow that escalated fast

[daniel habib @DannyHabibs]

Head-tracked “Window Mode.”

Your front camera finds your head. The view reprojects in real time so the screen feels like a window into the 3D scene.

True3D, no glasses.

[Aadit Sheth @aaditsh]

The best visual guide to LLMs you’ll ever see

[alphaXiv @askalphaxiv]

First paper published by Meta Superintelligence Labs!

In this paper, they make RAG faster by swapping most retrieved tokens for precomputed & reusable chunk embeddings, called REFRAG

This method improves its speed by 30x and fitting 16x longer contexts without accuracy loss

[Sumanth @Sumanth_077]

I found a 100% open source multi-agent workforce that runs locally!

Eigent is a desktop application that lets you build, manage, and deploy custom agent workforces to turn complex workflows into automated tasks.

100% Open Source

[Femke Plantinga @femke_plantinga]

Is your RAG system a paperweight?

It is if it can't handle a simple follow-up question.

Building basic RAG is easy. The real challenge is engineering systems that go beyond simple retrieval and actually do more with your data. This is how you build a RAG pipeline that can think, reason, and adapt.

Here's how advanced 𝗽𝗿𝗼𝗺𝗽𝘁𝗶𝗻𝗴 𝘁𝗲𝗰𝗵𝗻𝗶𝗾𝘂𝗲𝘀 transform your RAG pipeline:

𝗖𝗵𝗮𝗶𝗻 𝗼𝗳 𝗧𝗵𝗼𝘂𝗴𝗵𝘁 (𝗖𝗼𝗧): Instead of jumping to an answer, the model breaks down complex queries into intermediate steps, making the LLM "show its work."

𝗧𝗿𝗲𝗲-𝗼𝗳-𝗧𝗵𝗼𝘂𝗴𝗵𝘁𝘀 (𝗧𝗼𝗧): Takes reasoning further by exploring multiple paths simultaneously. The system generates several potential solutions and evaluates which is most promising. This is how you systematically weigh different pieces of evidence from multiple retrieved documents.

𝗥𝗲𝗔𝗰𝘁 (𝗥𝗲𝗮𝘀𝗼𝗻𝗶𝗻𝗴 + 𝗔𝗰𝘁𝗶𝗻𝗴): This framework lets the system *think* and *act* dynamically. It can reason about what information it needs, act to retrieve it, and then reason again based on what it found.

But beyond prompting, you need to fix your retrieval, too.

𝗤𝘂𝗲𝗿𝘆 𝗥𝗲𝘄𝗿𝗶𝘁𝗶𝗻𝗴 & 𝗘𝘅𝗽𝗮𝗻𝘀𝗶𝗼𝗻: Before hitting your vector database, change that vague user question into something your retrieval system can actually work with. This isn't just about synonyms, it's about understanding intent.

𝗔𝗱𝘃𝗮𝗻𝗰𝗲𝗱 𝗥𝗲𝘁𝗿𝗶𝗲𝘃𝗮𝗹 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗲𝘀: Move beyond basic search. Think hybrid search (combining similarity and keyword), metadata filtering, and multi-step retrieval for complex queries.

This is where it really gets interesting: 𝗔𝗴𝗲𝗻𝘁𝗶𝗰 𝗥𝗔𝗚.

AI agents can reformulate queries on the fly, re-retrieve information if initial results miss the mark, and handle queries requiring multi-step reasoning across multiple documents. Moving from a one-shot solution with no reasoning into an intelligent system that thinks, reasons, and adapts.

Ready to go from basic RAG to a reasoning engine? We cover these techniques and more in our ebook on Advanced RAG Pipelines.

𝗗𝗼𝘄𝗻𝗹𝗼𝗮𝗱 𝘆𝗼𝘂𝗿 𝗳𝗿𝗲𝗲 𝗰𝗼𝗽𝘆 𝗵𝗲𝗿𝗲: https://t.co/X6VZdOVUgc

[Evan @StockMKTNewz]

Here’s how Americans 🇺🇸 spend their day

[Darshak Rana ⚡️ @thedarshakrana]

But here’s where Wigner went too far

He said:
The collapse doesn’t happen just because something *interacts* with a particle

It only happens when a *conscious mind* observes it

In his words:
"It is not possible to formulate the laws of quantum mechanics without reference to consciousness"