Mike Elgan

How one man’s vibe-coding session evolved into a reckless global experiment where nobody’s accountable. Read my opinion column at Computerworld.com.

First they came for your eyes. Now they want your heart. Read my opinion column at Computerworld.com.

As we slouch toward a world of AI-made fake everything (and the distrust that follows) it’s time to spell out exactly how you’re using AI. Read my opinion column at Computerworld.com.

(Disclosure: The words and image in this post were made without AI.)

Superintelligent hosts Mike Elgan and Emily Forlini explore the link between churning homemade butter and the craving for authenticity in the new world of fake AI everything. They talk about Google’s new Gmail AI features as the most recent example of companies pushing everyone to use AI for communication. The hosts allow themselves to be influenced by influencer and Instagram chief Adam Mosseri, who says that in 2026 influencers will need to prove their humanity by posting authentic content and disclosing how they use AI.

Subscribe by searching your podcast app for "Superintelligent"

Perfection and polish now signal cheap AI. To communicate with authenticity, you’ll need to show up as yourself. Read my column at Computerworld.com.

Chinese parents using AI to shame their adult children into marriage aren’t weird. They’re just ahead of the curve. Read my column at Machine Society.

• Maintain a noble character and control your temper.

• Embody modesty and manliness.

• Practice piety, generosity, and simplicity in diet while avoiding the habits of the rich.

• Invest liberally in good education and teachers.

• Endure pain, work hard, desire little, and ignore slander and partisan games.

• Avoid superstition and empty enthusiasms, tolerate plain speaking, and embrace the discipline of philosophy.

• Reform one's character, avoid ostentatious rhetoric, and read texts with deep precision rather than superficiality.

• Exercise moral freedom, rely solely on reason, and remain consistent in behavior regardless of fortune.

• Live according to nature with dignity, graciousness, and affection, while remaining free from passion.

• Correct others gently by setting a good example rather than by criticism or fault-finding.

• Recognize and avoid the malice, caprice, and hypocrisy often associated with absolute power.

• Fulfill social obligations and avoid using "busyness" as an excuse to neglect duties to others.

• Accept criticism from friends even when unreasonable, praise teachers heartily, and genuinely love children.

• Love family, truth, and justice, and cherish the ideal of a commonwealth based on equity and freedom of speech.

• Exhibit self-mastery, cheerfulness in illness, and genuine integrity without the need for display.

• Act with unshakeable resolution, suppress vanity, work diligently, and treat all with impartial justice and temperance.

• Be grateful for good relationships, the preservation of health, and the guidance to live in obedience to nature.

Most people now get their news from social sites — and that’s a big problem. Here’s the solution. (Read my opinion column at Computerworld.com.)

Using digital twin technology to simulate real people can be powerful. But when it’s used to represent the views of people (leaders, celebrities, customers, or the public), people tend to oppose the idea. When they interact with prominent people, they want to interact with the real people. When their views are considered, they want to be asked rather than represented by a digital twin. 

This might change later, but for now the lesson is clear: it’s important to use digital twin tech carefully — and watch out for reputation harm from using it wrongly. Read my opinion column at Computerworld.com.

Host Leo Laporte is joined by Daniel Rubino, Sam Abuelsamid, and me! We do a post-mortem on Black Friday. We dig into record online sales, shaky in store traffic, and the tricks retailers use to turn inflated prices and AI shopping tools into “record breaking” headlines. From defense tech stumbles at Anduril to China leapfrogging the US on “open” AI models, the we point out how a shortage of shame runs through everything from autonomous weapons to space data center fantasies and EV pullbacks.​

You will also hear our honest takes on Trump’s Silicon Valley powered AI agenda, the Warner Music and Suno AI music deal, leaks about ads coming to ChatGPT, and rumors of a Jony Ive and Sam Altman hardware device, plus why Roblox, spying cars, and out of control RAM prices all point to the same broken incentives. 

This Week in Tech records live every Sunday at 5:15 p.m. Eastern, 2:15 p.m. Pacific, 22:15 UTC, and you can watch or subscribe on Apple Podcasts, Pocket Casts, RSS, YouTube, or through Club TWiT.

Check out the latest episode of Intelligent Machines with Leo Laporte, Paris Martineau, and me, Mike Elgan. 

In this episode, we push the conversation to the edge: Could the future of AI be sitting right on your desk instead of hiding in giant data centers? We dig into Nvidia’s new Blackwell chip and HP’s ZGX Nano—tiny but powerful machines that could let people fine-tune advanced AI models at home. 

Other stories include ChatGPT Atlas, Claude Skills, new research on AI news mistakes, OpenAI’s crackdown on Sora 2 deepfakes, and a look at why some experts say AI hype is getting out of hand.

Go here to listen, watch, subscribe and join the club!

This interview, conducted by Bill Kennedy, was surprisingly detailed, candid and biographical. Check it out! Also: Go here to subscribe to the Ardan Labs podcast. 

American federal agencies and private companies are now allowing weaponized foreign disinformation to spread and grow. Read my newsletter post at Machine Society.

Everyone should be clamoring for tools and features that block AI content. And for content companies, blocking AI slop is your new Job One. Read my opinion column at Computerworld.com.

In this episode of Superintelligent, we dove headlong into the spectacle of the Meta Connect event, where Mark Zuckerberg and team set out to convince us that the future is wearable, voice-activated, wrist-twitching and dorky.

Yes, we’re talking about the Meta Ray-Ban Display Glasses and their neural band companion—the gadget duo supposedly destined to replace keyboards and smartphones.

The demo was a train wreck. The AI hallucinated. Mark couldn’t answer a phone call. But it’s also true that a bad demo doesn’t mean it’s a bad product.

Meta’s glasses raise big questions about social acceptability, privacy, and whether the market will truly take off like the Silicon Valley giants hope it will. Also: Can Apple’s design sophistication rescue wearable tech from utter dweebdom? We pitted optimism against skepticism, debating everything from the speed of miniaturization to necessity of face computers.

Watch, listen and subscribe!

This episode of Superintelligent urgently explores the impact of social platforms on our daily lives. We opened against the backdrop of a tragedy — the assassination of political influencer Charlie Kirk. Watch, listen and subscribe on the Superintelligent website.

Yes, it is. So, what are the implications of these changes? Read my opinion column at Computerworld.com.

This week on Superintelligent, we tackled the spectacular rise of “Blame AI!” culture, where everyone from politicians to pop stars can now attribute any inconvenient fact, viral video, or even White House trash-bag scandal to AI-generated fakery. We explored how ordinary cultural events—like a K-pop musical hit or a Will Smith Instagram—that seem “too perfect” now get hit with AI accusations, often as a kind of backhanded compliment. Watch or listen to the Superintelligent podcast.

AI wearable startups: The future is glasses and watches, not pendants and pins. Read my opinion column at MachineSociety.ai.

Fake AI images can lie. But people lie, too, saying real images are fake. Call it the ‘liar’s dividend.’ Call it a crisis of confidence. Read my column at Computerworld.com.

AI reduces the learning curve and time frame for cyberattacks. Here comes the vibe hacking revolution. Read my opinion column at Computerworld.com.

The leading lifelogging expert once told me it was impossible — until we got the right AI. Now, we’ve got it. Read my opinion column at Computerworld.com.

New research shows how easily manipulated people are by AI tools, especially when they're designed to manipulate. Knowledge is the only defense. Read my opinion column at Computerworld.com.

People are getting answers from AI chatbots instead of the websites that feed those chatbots. Will that kill content sites? I say: Probably not. Read why on my Machine Society newsletter. 

This episode of the Superintelligent podcast covers a range of timely tech topics, including the recent launch and user reactions to GPT-5, broader trends in AI development, and the shifting landscape of platform monetization and product quality. Mike and Emily explore the role of AI in medical research, the complexities of lifelogging and information retrieval, and the evolution of note-taking tools from analog methods to modern AI-powered apps like MyMind. They also discuss the features and challenges of e-readers and tablets for digital annotation, as well as practical workflow tips for organizing travel and productivity.

Subscribe free by searching for "Superintelligent" in your podcast app, or subscribing on the Superintelligent webpage.

In this thrilling episode of Superintelligent, Emily and I talk about GPT-5 and how it performed worse than GPT-4 on everyday tasks, how AI monetization is changing how information is surfaced and prioritized, how new AI browsers and features like Google’s Overviews increasingly extract content and disrupt traditional web traffic, and the new trend of schools are rapidly banning phones to combat distraction and restore focus, reflecting a wider cultural struggle over attention and digital habits. Watch, listen and subscribe at our Substack.

Paywalled content is no match for the new generation of AI browsers, which instantly give users all the information in a paywalled story for free. Read all about it at MachineSociety.ai.