AI News Roundup: Jul 5, 2026

This week:

• The only AI glossary you’ll need this year
• The Companies That Built AI Are Now Using It to Cut Their Own Jobs
• DeepSeek Just Released Free AI That Rivals the World's Best. Again.
• AI Has Turned the Phone Into a Source of Fear for 82% of Americans
• AI Is Becoming America's Unofficial Doctor for the People Who Can't Afford One

The only AI glossary you’ll need this year

2026-07-04

Back when I was at Apple in the early days, we did not have terms like large language models or API endpoints. We just had engineers and customers trying to figure out how computers could actually help people. Today, the same gap exists between developers and regular folks navigating AI features in our phones, workplaces, and homes. TechCrunch just published a living glossary that finally bridges that divide by stripping away the hype and explaining what these tools actually do. They break down chain of thought reasoning, AI agents, and compute power into definitions you can use in a boardroom or at a family dinner. The jargon was never the point. The tools were. You just needed someone to translate them. Reading this guide feels like getting a reliable mentor instead of another sales pitch. It is optimistic without being naive and honest about what these systems can still not do. This guide respects your intelligence while acknowledging the real limits of today's technology. What term has tripped you up lately? Share it in the comments and let us demystify it together.

💡 Read my deep dive: Get Sirios: The Personal AI Product Apple Waited 50 Years to Deliver

Read the full story on techcrunch.com

The Companies That Built AI Are Now Using It to Cut Their Own Jobs

2026-04-27

Something just happened that I've been watching for thirty years and still didn't expect to come this fast.

Meta announced 8,000 job cuts this week, the largest single reduction in the company's history. They won't fill another 6,000 open positions. Microsoft offered voluntary buyouts to 8,750 employees. In a single week, two of the most valuable companies in the world signaled that roughly 20,000 human jobs are being redirected to machines. Both companies said the same thing: we're investing in AI, and AI changes what we need people to do.

I spent years at Cisco watching technology change job descriptions overnight. The video conferencing tools I helped build at WebEx eventually eliminated entire travel coordination departments. At Salesforce, the CRM software I worked on automated work that whole teams of people used to do manually. Every time, the fear was real. Every time, people adapted. But here's the part that's different now: the companies that BUILT the AI are the first ones using it to reduce their own workforce. The engineers who shipped the models are watching their employers say "we need fewer people now."

Goldman Sachs says AI is erasing 16,000 U.S. jobs per month. BCG just found that AI will reshape 50 to 55 percent of all jobs in the next three years. Not replace. Reshape. The distinction matters, but it doesn't make this week less real for 20,000 people.

Terrifying if you're waiting for things to go back to normal. Clarifying if you've learned that they never do.

The question worth sitting with: if the most skilled workers at the most valuable companies in the world can't protect their jobs through expertise alone, what skills actually matter now? And are you building them?

#AI #DigitalRenAIssance

Read the full story on cnbc.com

DeepSeek Just Released Free AI That Rivals the World's Best. Again.

2026-04-24

At Apple in 1997, when Steve Jobs came back and I was rehired to help turn the company around, we were ninety days from bankruptcy. One of the things that had nearly destroyed Apple was the PC clone wars. IBM opened its architecture, and dozens of companies made cheaper versions of what we had built over years. It felt like a threat. It turned out to be the event that put a computer in nearly every home in America.

DeepSeek released its fourth generation of AI models today. Free. Open source. Available to anyone at chat.deepseek.com. Running on Chinese chips that Washington tried to ban.

One year ago, this Chinese lab released a model that matched GPT-4 for almost nothing and Silicon Valley panicked. V4 is bigger and cheaper. The smaller version costs $0.14 per million tokens. GPT-5.5 charges $5.00 for the same task. Thirty-five times cheaper. Stanford's 2026 AI Index confirms Chinese AI has "effectively closed the performance gap" with US rivals.

Real tension here. For American companies that invested billions, watching a competitor match them for almost nothing is unsettling. The chip export controls were supposed to slow China's AI development. Did they? Or did they just change which chips got used?

The printing press did not ask which country invented movable type. It just spread. Every free, powerful AI model that enters the world means the Digital RenAIssance gets a little more democratic for regular people everywhere.

Does the source of your AI tools matter as much as the tools themselves? Or does free and powerful win every time?

#AI #DigitalRenAIssance

💡 Read my deep dive: Get Sirios: The Personal AI Product Apple Waited 50 Years to Deliver

📚 Explore more: The Digital RenAIssance

Read the full story on bloomberg.com

AI Has Turned the Phone Into a Source of Fear for 82% of Americans

2026-04-23

At Cisco, I built WebEx into a billion-dollar collaboration platform. The whole premise was simple: technology should make it easier for people to connect. I believed that then. I still do. Which is why data released this week stopped me cold.

A new Truecaller survey of more than 1,600 Americans found that 82% have ignored an important call or text in the past year because they feared it was a scam. In 2024, that number was 59%. That is a 23-point jump in a single year.

The telephone has been how human beings reached each other across distances for 150 years. It is now a source of dread. Seventy-five percent of respondents were targeted by a scam call or text in the past twelve months. One in four lost actual money. And 30% say they received a deepfake voice call that impersonated a family member, a celebrity, or a public figure. The kind that felt more convincing than any scam they had encountered before.

Here is the part that sticks with me. It takes just three seconds of audio to clone someone's voice. Three seconds. If you have a video on social media or a voicemail anywhere, that is enough. As one cybersecurity expert told InvestigateTV this week, his 8-year-old son can now make a deepfake.

Doctors are calling with test results. Schools are calling about kids. Clients are calling to close deals. People are missing all of it because they cannot tell what is real. Truecaller calls this "communication paralysis." That is exactly the right phrase for what AI has done to the most basic form of human connection we have.

I spent my career building tools that assumed the person on the other end of the call was who they said they were. That assumption has cracked. And carriers, government, and regulators do not appear to be moving fast enough to repair it. Seventy-five percent of those surveyed say the U.S. government is not adequately protecting consumers from AI-driven scams.

Who should be responsible for rebuilding trust in everyday communication? And what would it actually take?

#AI #DigitalRenAIssance

Read the full story on globenewswire.com

AI Is Becoming America's Unofficial Doctor for the People Who Can't Afford One

2026-04-22

At Alarm.com, I helped build sensors that detected when my father fell alone in his Florida home. The system called for help automatically. I wrote about that moment because it was the first time I understood AI in healthcare as something personal rather than theoretical.

A Gallup survey of more than 5,500 Americans shows how many people have their own version of that moment. One in four Americans has now used AI for health information or advice. That's more than 66 million people. But the number that stopped me: 14% of recent AI health users did so because they simply could not pay for a doctor visit. Among households earning less than $24,000 a year, that number climbs to 32%. Two out of three people in that income bracket turning to a free chatbot when something hurts or worries them, not out of preference but because better options aren't within reach.

The data gets more specific. Twenty-one percent used AI not because of cost or access, but because they felt dismissed or ignored by healthcare providers in the past. Not cost. Not geography. Dignity.

We debate whether AI will disrupt healthcare. For these 66 million Americans, the disruption already happened. Quietly. Through a free app on a phone, filling gaps the system left open. And almost half of those users say AI made them more confident when they did talk to a doctor.

This is what "AI for everyone" looks like when you zoom past the tech industry. Who should be responsible for closing these gaps: the companies that built the tools, or the system that made them necessary?

#AI #DigitalRenAIssance

Read the full story on news.gallup.com

Want these updates as they happen? Follow me on LinkedIn or X.