Microsoft Just Did What Google Couldn't: Built a Quantum Computer That Actually Works
4IR - Daily AI News
Welcome back to 4IR. Here's today's lineup:
Microsoft's quantum computer actually works - New chip solves the error problem that's killed quantum computing for 30 years
Cloudflare drops tools to catch rogue ChatGPT use - Companies can finally see which employees are feeding secrets to AI
Ex-Salesforce execs raise $82M to replace Salesforce - Two startups close massive rounds in plan to make CRM actually work
🔥 TOP STORY: Microsoft finally builds a quantum computer that doesn't break
The story: Microsoft just did something everyone said was impossible. Working with UC Santa Barbara physicists, they built a quantum computer that actually works reliably—the first one ever. Called Majorana 1, this chip fixes its own errors instead of breaking every five seconds like every other quantum computer. Unlike competitors who need thousands of backup systems to catch mistakes, Microsoft's approach spreads information around so errors can't kill it. Think of it like backing up your files automatically across multiple computers.
What we know:
First quantum computer that works reliably without constant errors
Based on physics theory from 1937 that nobody could build until now
Fixes its own mistakes without needing backup systems
Published in Nature, unveiled at Microsoft's Santa Barbara conference
Team led by UCSB Professor Chetan Nayak
Why it matters: Regular quantum computers are like Formula 1 cars—insanely powerful but break if you look at them wrong. Microsoft's chip is more like a Toyota—maybe not the fastest, but it actually works reliably. This could be the breakthrough that makes quantum computing practical for real businesses, not just research labs.
Microsoft took a different path than everyone else. While Google and IBM focused on raw power, Microsoft spent a decade on experimental physics most people dismissed. The payoff could be significant. This explains their confidence in promising quantum computing on Azure by 2029. The challenge is scaling from a small prototype to commercial deployment. If successful, they'll have a major advantage—but that's still a big if.
🏛️ POLICY: Cloudflare goes nuclear on Shadow AI
The story: Cloudflare kicked off AI Week 2025 with a reality check: your employees are feeding company secrets to ChatGPT and you have no clue. Their new security tools detect "Shadow AI"—when workers use AI behind IT's back. The platform blocks risky AI use, creates safe channels for approved tools, and stops your data from training the next ChatGPT. They're also helping websites block AI scrapers from stealing their content without permission.
What we know:
Detects when employees use unauthorized AI tools
Creates approved channels for safe AI use
Cuts AI costs by avoiding duplicate work
Blocks AI bots from stealing website content
Plugs into existing security systems
Why it matters: This is every company's worst nightmare—employees treating ChatGPT like Google while uploading customer lists and source code. Cloudflare's offering the corporate equivalent of parental controls. Companies can finally say yes to AI without losing control of their secrets.
Cloudflare's timing is excellent. "Shadow AI" clearly communicates the risk—employees using unapproved tech tools. Rather than blocking innovation, they're providing structure. Being both an AI provider and security company gives them unique visibility into emerging threats. Microsoft and Google will likely develop similar offerings soon, but Cloudflare has first-mover advantage in defining the category. Expect strong demand from enterprise security teams in the coming weeks.
💰 FUNDING: Salesforce alumni raise $82M to replace their old company
The story: Two AI startups just raised massive rounds to blow up the CRM market. Attio grabbed $52 million (Google Ventures leading) while Aurasell scored $30 million in a round that closed in just 28 hours. Both are building sales software from scratch with AI at the core, not added as an afterthought. Instead of juggling 15 different tools, they promise one platform that does everything. The speed of funding—especially Aurasell's overnight close—shows investors smell blood in the water around Salesforce's empire.
What we know:
Attio now valued over $500 million
Aurasell's funding closed in record 28 hours
Both founded by former Salesforce executives
Promise to replace dozens of separate sales tools
Users report working 3-5x faster
Why it matters: Salesforce is the new Oracle—bloated, expensive, and ripe for disruption. These startups aren't adding AI features to old software. They're building what sales teams would design if starting fresh today. The 28-hour funding says everything about investor confidence.
Both founders are Salesforce veterans who know exactly where the product falls short. They're not adding features to compete—they're rebuilding CRM from scratch for the AI era. That 28-hour funding close shows how desperate investors are to find the next big thing in enterprise software. Salesforce will probably acquire one of these companies or their competitors within the year. But with a $300 billion market cap and thousands of enterprise customers locked into contracts, they have plenty of time to respond. The real question is whether they can innovate fast enough internally or if they'll need to buy their way back to relevance.
⚡ QUICK HITS
Apple commits $600B to U.S. chip manufacturing - Massive expansion with Texas Instruments, GlobalFoundries for iPhone chips
GPT-5 hits 700M weekly users - Enterprise rollout begins as Amgen reports breakthrough results
GoodShip raises $25M Series B - Bessemer backs freight platform used by Tropicana and Kellanova
Eyebot secures $20M from General Catalyst - Vision care kiosks deliver prescriptions in 90 seconds
xAI privacy breach affects 370K users - Grok conversations exposed on search engines amid Apple lawsuit