💡 “No” Is Becoming AI’s Most Expensive Feature

In partnership with

☕ Morning! 💡 Your Weekly 5-Minutes of Caffeine and Tech Clarity

Quick Hits 🎯

• 🧭 Models are getting better at saying no, and teams are learning that “no” has a price tag

• 🤖 Google put agentic AI inside Gemini for businesses, so the demo era is officially over

• ☁️ Cloudflare acquired Deno to rebuild how code actually runs at the edge

• 🎛️ A non-text AI model called Jev hit a $7.5B valuation weeks after launch - on efficiency, not scale

• ⏸️ Skype’s founding engineer Jaan Tallinn is telling the industry to slow down, and he’s funding the argument

• 🎧 Spotify is open-sourcing its internal tooling and quietly selling software to enterprises

• 🎤 TechCrunch Disrupt 2026 is live. The founder’s field guide just dropped

🎁 + 4 other stories you might find useful

Our Partner 🎉 

Master Claude AI (Free Guide)

The professionals pulling ahead aren't working more. They're using Claude.

Our free guide will show you how to:

• Configure Claude to be the perfect assistant

• Master AI-powered content creation

• Transform complex data into actionable strategies

• Harness Claude’s full potential

Transform your workflow with AI and stay ahead of the curve with this comprehensive guide to using Claude at work.

Get The Free Guide

The Big Picture 🖼️

💡 Google Moved Agentic AI From the Demo Stage to the Org Chart.

Gemini can now run tasks across apps for business users, with identity and policy rules wrapped around it from the start.

Why it matters: most “agents” so far have been impressive on stage and untrustworthy in production.

Pairing autonomy with governance from day one is the move that makes an agent something a CISO will actually sign off on, and it forces every rival to match both at once, not one at a time.

The takeaway: autonomy without control was always the easy half. What Google is really selling here is the hard half.

💡 The Next Model War Is About Efficiency, Not Size.

Jev, a non-text AI model, pulled a $7.5B valuation weeks after launch and the story isn’t the number, it’s what investors paid for.

Why it matters: the bet is on latency, cost per token, and reliability across tasks, not parameter count.

For anyone budgeting AI spend, that’s the signal that the economics are tilting toward lean, fast runtimes you can actually afford to run at scale.

The takeaway: the moat is moving from “biggest” to “cheapest to run well.” The valuation math behind that shift is the part worth reading twice.

💡 Cloudflare Didn’t Buy a Company. It Bought a Runtime.

The Deno acquisition is aimed squarely at Workers. A faster, more secure execution model baked into the edge.

Why it matters: speed at scale comes from infrastructure choices, not clever code. Owning a modern runtime means predictable behavior across regions and teams, which is exactly what you need when “move fast” has to coexist with “don’t break compliance.”

The takeaway: the edge wars are now a runtime war. What this acquisition tells you about where serverless is headed matters even if you never touch Workers.

💡 The Durable AI Moat Is the Plumbing, Not the Model.

The throughline from Disrupt 2026: companies built on one clever model age badly. The ones that last are built on data flywheels and cross-tool workflows.

Why it matters: demos don’t compound, integration does. Governance, standards, and open ecosystems are turning into the real defensibility, because they’re the parts competitors can’t clone over a weekend.

The takeaway: if your advantage is a prompt, it’s already gone. Here’s the founder playbook that’s actually holding up.

💡 One of AI’s Earliest Funders Wants Everyone to Slow Down.

Jaan Tallinn, who bankrolled a lot of the field, is making the case that deliberate caution is a competitive strategy, not a brake.

Why it matters: slowing down thoughtfully buys trust, regulatory alignment, and the room to deploy in markets where one bad rollout ends the whole relationship. In reputation-sensitive sectors, that’s not conservatism. It’s survival.

The takeaway: the people closest to the money are getting nervous. What Tallinn is actually worried about is more specific than the headlines suggest.

💡 The Optimists and the Safety Hawks Are Starting to Agree.

A roundup of incremental wins across biology, tooling, and defense tech shows capability getting cheaper and more deployable, fast.

Why it matters: the practical read isn’t “build faster” or “build slower.” It’s build modular. Interoperable systems that absorb new capability without tearing up your governance every quarter.

The takeaway: speed and safety stop being opposites once your architecture is pluggable. The week’s case for cautious optimism lands on the same conclusion from the other direction.

💡 Spotify Is Turning Its Internal Tooling Into a Business.

Spotify is open-sourcing internal tech and getting serious about selling enterprise software.

Why it matters: when a consumer company’s tooling is good enough to sell, it signals where the margins are moving, and that platform and ecosystem plays beat point products. Watch who else starts monetizing the systems they built for themselves.

The takeaway: your internal platform might be a product you haven’t noticed yet. Why Spotify thinks enterprises will pay for it.

Trending Tools 📈

huggingface/transformers

(+72 ⭐ per day, 🐍 Python) Link

• The model-definition framework behind most state-of-the-art text, vision, audio, and multimodal work

• Helps teams: go from experiment to production-ready deployment without rebuilding the stack each time

NationalSecurityAgency/ghidra
(+88 ⭐ per day, ☕ Java) Link

• Full software reverse-engineering suite for analyzing binaries across platforms

• Helps teams: strengthen malware analysis, vulnerability research, and security audits with one extensible toolkit

f/****prompts.chat
(+33 ⭐ per day, 🟧 HTML) Link

• A community hub for sharing and curating prompts, self-hostable for privacy

• Helps teams: standardize prompt engineering so prototyping stops being a solo guessing game

projectdiscovery/nuclei
(+7 ⭐ per day, 🔵 Go) Link

• Fast, YAML-driven vulnerability scanner with a huge community ruleset

• Helps teams: automate vuln discovery across apps, APIs, networks, and cloud configs

Project-HAMi/HAMi
(+5 ⭐ per day, 🔵 Go) Link

• Heterogeneous GPU sharing on Kubernetes

• Helps teams: squeeze more out of multi-tenant GPU clusters and cut idle compute cost

openbq-org/OpenBB
(+5 ⭐ per day, 🐍 Python) Link

• Open data platform for analysts, quants, and AI agents

• Helps teams: centralize data access and analytics in one extensible layer agents can plug into

oven-sh/bun
(+4 ⭐ per day, 🟨 Zig) Link

• JavaScript runtime, bundler, test runner, and package manager in one

• Helps teams: collapse the toolchain and speed up dev and CI pipelines

ClickHouse/ClickHouse
(+2 ⭐ per day, 🔷 C++) Link

• Real-time OLAP database built for fast, large-scale analytics

• Helps teams: power dashboards and observability with low-latency queries at scale

Tech Trend of The Week 📊 

🔍 Searches for “agentic AI” edged up 4% over last week.

It’s a modest move, not a spike - and that’s the more interesting signal. After Google wired agents into Gemini for business, interest isn’t surging and crashing. It’s holding at a steady, elevated baseline.

The signal: “agentic AI” is graduating from hype term to default vocabulary. Words that plateau high don’t fade. They get absorbed into how the whole industry talks. The spikes are behind us; the adoption is ahead.

Our Partner 🎉 

Stop Drowning In AI Information Overload

Your inbox is flooded with newsletters. Your feed is chaos. Somewhere in that noise are the insights that could transform your work—but who has time to find them?

The Deep View solves this. We read everything, analyze what matters, and deliver only the intelligence you need. No duplicate stories, no filler content, no wasted time. Just the essential AI developments that impact your industry, explained clearly and concisely.

Replace hours of scattered reading with five focused minutes. While others scramble to keep up, you'll stay ahead of developments that matter. 600,000+ professionals at top companies have already made this switch.

Join them today, for free.

Hit Reply And Tell Me 💬 

What's the most impressive tech you've seen recently that actually works?

I read every single reply

Go where your customers actually are 📍

Most of the internet copies Reddit.

Blogs, SEO pages, AI models, product reviews. They all pull from the same well.

Reddit is the 3rd biggest site on earth. And it trains the bots your users ask for help.

Odd Angles Media runs Reddit campaigns for over 45 brands each month.

Luckily, we’ve convinced them to give away all of their strategies in a (free) ebook ⬇️⬇️

Grab The Free Ebook (Free) →

Reach the People Who Sign the Checks 💰️ 

40,000+ CTOs and engineering leaders. 96% US-based. 80%+ corporate emails.

They evaluate vendors. They shortlist solutions. They buy.

Advertise with us on Paved → (discount applied)