AI News Roundup: Jul 13, 2026

This week:

• Waze is getting a bunch of new AI-powered features
• OpenAI bets on families as ChatGPT goes deeper into households
• Artificial intelligence: University of Chicago Law School to ban phones, laptops in classroom for 1st-year students in AI strategy - ABC7 Chicago
• Meta removes controversial AI feature on Instagram after backlash
• OpenAI releases new voice models for more natural live conversations
• Google Photos adds a new AI ‘Video Remix’ tool
• Amazon will stop accepting new customers for Mechanical Turk
• The ‘first’ AI-run ransomware attack still needed a human
• Savi’s app aims to protect consumers from realistic AI scams like kidnappers demanding ransom
• Police use of artificial intelligence grows as rules lag behind - Journal-News.com

Waze is getting a bunch of new AI-powered features

2026-07-13

Waze just folded conversational AI into the driving app so you can report traffic and find destinations with your voice. You say “find a coffee shop open right now” and it figures out the route. They also added a less chatty mode so directions stop talking over your podcast. It looks like another step toward AI that actually fits into daily life instead of demanding your full attention. In thirty years building products at Apple, Cisco, and Salesforce, I’ve seen plenty of tech promises that never reached the dashboard. This one might finally stick because it solves a real friction point: drivers don’t want another screen to stare at. They just want to know where the potholes are and how to avoid the backup without missing their turn. People worried GPS beepers would erase our sense of direction back in the nineties. They didn’t. We just learned to read the map and trust the voice. This feels like that same quiet shift, only faster. Think of it like a seasoned local driver riding shotgun. You don’t need to program the exact turns. You just tell them you’re running late and they know which side streets cut through the construction. The training happens once in quiet data centers. You use it every morning while stuck behind road work. It is the AI for the rest of us model in action, getting you to your destination with less mental load. The convenience is real, even if handing over the steering logic to a car screen feels like stepping off a familiar ledge. How are you testing voice commands in your car yet?

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

Read the full story on theverge.com

OpenAI bets on families as ChatGPT goes deeper into households

2026-07-12

The day my dad fell in Florida, Alexa caught it and called for help. That moment proved smart home tech works best when it serves the whole house, not just one person.

OpenAI is finally catching on. Hiring a dedicated product manager for families and caregivers signals that AI is shifting from solo productivity tools to shared household utilities.

Remember when home security moved from professional monitoring boxes to a single camera on the porch? The pattern is familiar. Technology always starts as a gadget for the hobbyist and ends up running the living room.

Trusting AI in a family setting is like buying a house alarm. You do not worry about how the wiring works. You just want to know it will wake you up if something goes wrong, and keep the kids from accidentally triggering false alarms.

I see this playing out in quiet ways. A caregiver getting medication reminders without scrolling through apps. Parents setting shared household memory so the AI remembers key dates and school schedules. Safer guardrails for teenagers that actually work instead of just blocking keywords.

It looks like a convenience feature, but it is really a test of whether we will let machines manage our most private moments or keep humans firmly in the loop.

The Digital RenAIssance is not about replacing family time with smarter gadgets. It is about handing back the busywork so we can actually be present. What household chore are you hoping AI will take over next?

📚 Explore more: The Digital RenAIssance

Read the full story on techcrunch.com

Artificial intelligence: University of Chicago Law School to ban phones, laptops in classroom for 1st-year students in AI strategy - ABC7 Chicago

2026-07-10

The University of Chicago Law School just banned phones and laptops for first-year students.

It looks like a hard line drawn in the sand, but it is really a lesson in attention. We spent decades building devices that promised to free our minds. Now we are trapped by the notifications they generate.

I watched this pattern before. When calculators arrived in the 1970s, teachers feared we would forget long division. We learned to trust the multiplication and check the mortgage instead. The tool did not replace the math.

It just changed where we spent our mental energy. Think of a fresh restaurant kitchen. You train the head chef once on the core techniques, then you let them handle the nightly orders without constantly checking their phone.

The training happens in quiet focus, not in fragmented glances at a screen. Law students need to build the deep reading habit before AI can summarize it for them. If we outsource our concentration to a device, we lose the ability to hold complex ideas in our heads.

It is easy to worry this policy will feel outdated by graduation day. Terrifying if you have been coasting on digital multitasking. Liberating if you actually want to master your craft.

The bottleneck is no longer finding information. It is deciding what deserves our full attention.

How are you protecting your team’s focus time in a world that constantly interrupts?

Read the full story on news.google.com

Meta removes controversial AI feature on Instagram after backlash

2026-07-11

In thirty years building products at Apple and Cisco, I watched tech companies ship fast and ask questions later. Meta pulled an AI photo tool after three days of backlash, calling it a mistake in judgment rather than engineering. They skipped a basic human rule we have lived by for generations: you do not borrow someone else’s face without asking.

I watched this pattern repeat in the early days of internet file sharing. We built tools because we could, not because we should. The industry learned the hard way that speed without consent creates friction. Today’s AI tools are no different.

Imagine a kitchen table where someone starts redrawing your family photos without permission. You would ask them to stop. Tech companies are just learning how to handle that conversation at scale.

The tension here is simple but heavy. We are building machines that can mimic anything, yet we still need to teach them the difference between access and ownership. Terrifying if you think AI replaces your privacy. Liberating if we finally design it to respect the people behind the pixels.

How are you setting boundaries around personal photos and AI tools in your own life?

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

Read the full story on techcrunch.com

OpenAI releases new voice models for more natural live conversations

2026-07-08

When I built home safety tech at Alarm.com, the goal was simple: help when you can't ask for it. My dad fell in Florida. Alexa heard him before he could dial. That moment taught me voice isn't just convenience. It is safety.

Now OpenAI says voice is becoming the primary interface for work. GPT-Live-1 lets you interrupt and keep talking naturally. It handles turn-taking like a human friend, not a robot waiting for silence. The new models can listen while they speak, which changes everything for multitasking parents or busy executives.

My mom worked at an answering service before the digital age. She knew that listening was just as important as speaking. This new model feels closer to that human rhythm than the stiff commands of yesterday.

Think of it like upgrading from a calculator to a tutor. You still do the math, but now you can ask questions while calculating.

It is liberating to talk without typing. Terrifying when the machine knows too much about what you are thinking before you say it.

The Digital RenAIssance isn't about faster typing. It is about natural conversation for the rest of us.

How are you using voice assistants in your daily work? I want to know what is actually helping.

📚 Explore more: The Digital RenAIssance

Read the full story on techcrunch.com

Google Photos adds a new AI ‘Video Remix’ tool

2026-07-09

I spent thirty years in tech building tools at places like Apple and Cisco. People once feared cameras would ruin the art of remembering. They just changed the medium. Now Google Photos is adding Video Remix, and it feels like the next step in that journey. You can relight a dark clip or paint a video with watercolor strokes using just a tap. It removes the friction between your memory and sharing it. This is what AI for the rest of us looks like on a Tuesday afternoon. No complex timelines or render bars. Just you and your footage. Think of it like a restaurant chef plating the food you ordered. They do the heavy lifting on presentation, so you enjoy the meal. It follows the 80/5 rule perfectly: AI gets the first draft done in five percent of the time. You do the rest to make it yours. There is a tension here worth holding. The bottleneck is no longer Can I do this? It is Is this worth doing? A better question. A more human question. We are moving from making the tech work to deciding if it matters. How are you using these new editing tools in your own life?

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

Read the full story on techcrunch.com

Amazon will stop accepting new customers for Mechanical Turk

2026-07-07

Amazon is closing the doors to Mechanical Turk. The crowdsourcing service that quietly taught early AI models how to read, count, and understand human language is winding down. Think of it like a restaurant that finally gets its recipe right. You stop calling in outside cooks when the head chef can handle the whole menu. MTurk was that temporary kitchen for a generation of data labelers and sentiment taggers. I spent years building product teams at Cisco and Salesforce where we constantly needed human feedback to train early systems. We relied on those same crowdsourced workers to catch obvious mistakes before they reached everyday users. We faced similar scares when calculators replaced abacuses and voice recognition replaced switchboard operators. It feels like an ending for the people who built their income on micro-tasks, but it is also the moment AI finally grows up and learns to do its own homework. The bottleneck looks like it is no longer finding people to label images. It is whether the model actually understands what it sees. We get to stop paying pennies for synthetic intelligence and start demanding tools that think independently. If you used MTurk to train or test early AI tools, what was the one pattern you noticed that machines never quite got right?

Read the full story on techcrunch.com

The ‘first’ AI-run ransomware attack still needed a human

2026-07-07

The headline says an AI ran a ransomware attack from start to finish. That sounds terrifying until you read the fine print.

A human still picked the victim, rented the servers, and handed over the stolen passwords. The machine just turned the crank.

This shift looks familiar if you have thirty years in tech. We used to worry about the craft itself. Now we worry about who is managing the workflow.

Think of this AI agent like a high-speed intern on an assembly line. You still need someone to pick the raw materials and set up the factory floor. The intern just moves the boxes incredibly fast.

The real lesson here is about our daily habits. We treat passwords like a single front door key that disappears in a drawer. That old model is gone.

You now need to guard a digital fingerprint that never changes and follows you across every login screen. The bottleneck is no longer writing malware. It’s picking a victim and buying a server.

This means simple hygiene matters more than ever. Enable two factor authentication on your email and bank account today. It takes three minutes and stops ninety percent of automated break-ins before they start.

The machine handles the technical grind now, but humans still keep their hands on the steering wheel.

How are you locking down your accounts this week?

Read the full story on techcrunch.com

Savi’s app aims to protect consumers from realistic AI scams like kidnappers demanding ransom

2026-07-07

My mom worked at an answering service back when a wrong number was just a nuisance. Today it means an AI voice clone of your daughter screaming for help in a Walmart parking lot. That exact scenario just happened to the founders of Savi Security, and their response shows how fast technology moved ahead of our common sense.

The barrier to running a convincing kidnapping scam has collapsed. Three seconds of social media audio is all it takes to clone a voice. The FTC reports consumers lost three and a half billion dollars to impersonation scams last year. We used to keep spare keys under the mat. Now we are handing our families’ voices to algorithms that learn faster than we can warn them.

Savi launches a real-time call screening app this Tuesday. You add the app as a listener during a suspicious conversation, and it flags behavioral tells while you are actually on the line. It costs less than a streaming subscription for an entire family plan. Looks like we finally get a digital seatbelt for the phone in our pockets.

The fear is how easy this became, but the possibility is we finally get protection for every phone in our pockets. The bottleneck is no longer whether we can build AI safety nets. It is simply remembering to turn them on before the first ring goes out.

What’s the last time a suspicious call made you actually hang up and verify? I want to hear how your family is handling the new noise.

Read the full story on techcrunch.com

Police use of artificial intelligence grows as rules lag behind - Journal-News.com

2026-07-06

I want to start with a confession. For thirty years building tech at Apple, Cisco, and Salesforce, I watched every new tool move faster than the rulebook. We had the same panic with early email filters, then social media feeds, and now AI in law enforcement. History does not repeat itself exactly, but it sure rhymes.

The problem is not that police are using AI. It is that we are letting the tools write their own operating manual while everyone watches from the sidelines. Think of AI in policing like a new GPS system. It gets you to your destination faster, but if the map has outdated streets or wrong turns, you end up in the wrong neighborhood. We need to know who is drawing those maps and how they fix them when they break.

I have seen this pattern before with early biometric scanners and automated traffic cameras. The first deployments always look like magic until a single false positive changes someone’s life. That is why we need plain rules that require human review, transparent audits, and clear accountability before any algorithm touches a case file.

Terrifying if you have been waiting for perfect systems. Liberating if we actually build them with everyday people at the table.

How are you seeing local departments handle this right now? What guardrails do you think actually work in practice?

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

Read the full story on news.google.com

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