Brad Vogel

🧭 2) DeSantis Emerging as a Political AI Skeptic Recent news : including from Politico, frames Florida Gov. Ron DeSantis not just as a critic of big tech, but as a growing skeptic of AI and its infrastructure needs: • A headline from Politico states: “We have to reject that with every fiber of our being”: DeSantis emerges as a chief AI skeptic. This reflects a broader political narrative, not just grid concerns: skepticism about AI’s societal impacts, which too often gets tethered to infrastructure debates like data centers.

2

Dashboards don’t create accountability. They create meetings. Charts get shared. Everyone nods. Action items get “noted”. Same problems. Next quarter. Accountability doesn’t come from visibility. It comes from systems — where ownership is clear and decisions can’t be ignored. I wrote a short note on why dashboards feel productive but rarely change behaviour — and what actually does. If your org has great dashboards but the same issues every review… this will sound familiar. Dashboards Don’t Create Accountability — Systems Do: https://lnkd.in/gCGpm-VU

7

1 Comment

The The Wall Street Journal ran a deepfake voice quiz this week, according to an article written by Nicole Nguyen, Audrey Valbuena and Juanje Gomez Hinestrosa Players were given five audio clips - some real WSJ reporters, some AI-generated clones. The challenge was simple. Can you tell which voices are synthetic? Most humans couldn’t tell artificial from real. So we tested our detection systems on this challenge. It achieved a perfect score in less than a minute. The WSJ used OpenAudio software to clone their reporters' voices from just 10-30 seconds of publicly available social media clips. The results were convincing enough that most humans struggled to identify the fakes consistently. But the deterministic signal processing we use doesn't guess. It analyzes the acoustic characteristics of a voice at frame rates that cannot be processed by the human ear, but require minimal compute when processed by a modern CPU. This quiz may seem innocuous, but it comes at an interesting inflection point in history. CrowdStrike says voice clone scams surged 442% between the first and second half of last year. Criminals use this technology to defraud victims out of money and passwords. Senator Amy Klobuchar just spent hours trying to get a viral deepfake of herself removed from X, where a fake video showing her making vulgar comments accumulated over a million views. The platform refused to take it down or even label it as synthetic. Secretary of State Marco Rubio got impersonated by AI-generated voice and text messages that fooled high-level government officials including foreign ministers and a U.S. governor. So naturally, when human judgment becomes worthless, you need deterministic verification. Running in a closed loop on your mobile device or on cloud, our technology can detect synthetic audio in less than 100 milliseconds and scale to millions of concurrent streams. No machine learning that can be fooled by the next generation of synthesis models. No open circuits that create security risks. And no lapses. This isn't about being alarmist. The WSJ reporters used easily accessible software that runs on a laptop. It’s the same technology which is democratized and getting better fast.   In short, voice cloning will only become cheaper, faster, and more convincing. With UnCognito, You Can Trust What You Hear. #WSJ #TrustWhatYouHear #UnCognito #Authenticity #AI #Deepfakes #ElevenLabs #OpenAudio

32

6 Comments

I added an AI assistant to ADU Pulse. You can now compare towns, check local bylaws against state law, and see which provisions the AG has struck down - pretty excited about this feature. Ask it anything: "Compare ADU permits between Plymouth and Falmouth." It pulls from real EOHLC survey data, Census building permits, and bylaw analysis across 293 towns. It's not perfect yet - I'm building this in public and improving it every week. If something looks off or you have ideas, I want to hear it. Try it: adupulse.com

11

3 Comments

More and more members of congress are openly discussing how AI may soon threaten human control and human survival. This gives me hope at a time when hope is badly needed. I had the opportunity to speak with Bernie Sanders along with Eliezer Yudkowsky, Nate Soares, and Daniel Kokotajlo. We discussed how AI companies are racing to build AIs far smarter than humans, and how even company leaders admit these AIs will endanger humanity. Link to the video in the comments.

54

3 Comments

Claude Code is helping government too. I was amazed to hear from two software engineers inside government IT departments last week: - "Claude Code makes my 20 engineers feel like 200" and - "A project that used to take 1 month took just 1 weekend with Claude Code." Government software engineers build mission-critical systems affecting millions of residents, yet they're stretched thin maintaining legacy systems. Empowering them means better public services built in-house. Safe AI models and responsible deployment also matter even more here given the high stakes of government production code.

35

Writer Asa Fitch: "There is a decent chance that it all works out. "But it is at least equally likely that many of today’s massive contracts get postponed or renegotiated because end demand for AI services doesn’t grow in line with the infrastructure build-out." WSJ - Debt Is Fueling the Next Wave of the AI Boom -Today https://lnkd.in/eVunpc3q?

1