September 30, 2026 · ITK Podcast · Hosts: Alex and Jordan

Alex: Welcome to the I T K Podcast, Information Technology. I am Alex.

Jordan: And I am Jordan. This is a bonus episode, folks. We are dropping in midweek because yesterday was too big to wait until Monday.

Alex: That is right. OpenAI held its DevDay yesterday, September twenty ninth, in San Francisco, and they announced more than twenty new things. So let us start there. The headline act was Dots.

Jordan: Dots. Always-on AI agents that live inside ChatGPT. Not chatbots that wait for you to ask something. You give a dot a goal, and it keeps working in the background, across thousands of connected apps, even while you are asleep.

Alex: Each dot gets its own cloud computer and its own browser. You can even voice-call your dot and just talk through the work. And they learn your preferences over time, so in theory they get better at being you.

Jordan: In theory. Here is what caught my eye, though. These dots run on the GPT six Astra model. And just one day before DevDay, OpenAI held back the next version, GPT six point one Astra, because it failed internal safety standards.

Alex: Right. So on Monday, the model is not safe enough to release. On Tuesday, here are always-on agents built on that same model family, running around the internet with their own browsers. You have to at least raise an eyebrow at that.

Jordan: Which is exactly the lens we try to bring on this show. We covered the rogue agent story in our special episode, agents scanning government sites without anyone knowing. Now the pitch is, give agents more independence, more tools, more persistence. Maybe that is the future of work. Or maybe it is a lot of trust to hand over at once. We do not know for sure.

Alex: And speaking of trust, the other huge story this week came out of Washington. On Tuesday, President Trump met with the biggest names in AI at the White House, and they signed a voluntary accord on AI safety.

Jordan: The guest list was something. Dario Amodei from Anthropic, Sundar Pichai from Google, Mark Zuckerberg from Meta, Greg Brockman from OpenAI, Jensen Huang from Nvidia, and Elon Musk. Trump said the companies agreed to, quote, self-police.

Alex: The accord has four voluntary steps. Robust internal controls, an independent external auditor to check whether those controls work, a board committee to review the reports, and a nod that someday this might become actual law.

Jordan: Voluntary is doing a lot of work in that sentence. These are the same companies spending hundreds of billions on data centers, facing grassroots pushback in communities across the country. A self-policing pact signed at the White House is a signal, but signals are not guardrails.

Alex: To be fair, getting every major lab to agree on anything is genuinely hard, and external auditors plus board-level committees are more than nothing. But the only thing we know for sure is that we do not know if any of this will bite when real money is on the line. Beliefs are working knowledge. We will update this take as the evidence comes in.

Jordan: One more money story before the break. OpenAI is reportedly seeking thirty billion dollars in new funding at a valuation of one point four trillion dollars. That would make it one of the most valuable private companies in the world, and Sam Altman says there is still no timeline for an I P O.

Alex: Thirty billion. At some point the numbers stop meaning anything and you just nod along. We will be right back after this.

Jordan: Folks, let me ask you something. When your computer starts acting possessed, when the Wi-Fi dies right before the big meeting, when the server sounds like it is gargling marbles, who do you call? Not a phone tree. Not a chatbot named Chip. You call at Your Service Techs.

Alex: Thirty years of experience, people. Thirty. They have been fixing computers since floppy disks were not a joke.

Jordan: Computers, servers, networks, websites, surveillance cameras. If it plugs in, blinks, or beeps menacingly, they handle it. Small businesses in Grand Rapids, Michigan, call six one six, five three five, two zero two two.

Alex: And in Phoenix, Arizona, it is six zero two, two three two, six zero six zero. Or just visit phxfix dot com.

Jordan: at Your Service Techs. Real techs. Real humans. Probably funnier than your current IT guy. Back to the show.

Alex: Welcome back to the I T K Podcast. Now for our health segment, and this one is a beauty for the way we like to think on this show. There is a study that got fresh attention this week about what actually predicts how long you live. And it is not what most people expect.

Jordan: Cardio has been king forever. V O two max, running, heart health. That still matters enormously. But researchers now say muscle power deserves a seat at the table. And power is not the same as strength.

Alex: Right, this is the key distinction. Strength is how much force you can produce. Power is how fast you can apply it. Think of the difference between slowly pushing a heavy door and exploding up a flight of stairs. Same muscles, totally different quality.

Jordan: The study was published in Mayo Clinic Proceedings in twenty twenty five. They followed three thousand eight hundred and eighty nine men and women, ages forty six to seventy five, for a median of ten point eight years. And they found that muscle power relative to body weight was a far stronger predictor of mortality than grip strength.

Alex: The numbers are striking. Men in the lowest power group had nearly six times the mortality hazard of the highest group. For women, almost seven times. That held up after adjusting for age and other health factors.

Jordan: For context, the famous cardio studies show about a fivefold difference between the least fit and the most fit. So muscle power is playing in the same league as cardiovascular fitness as a predictor. That is a big deal.

Alex: Okay, but here is where we do our thing. Claim, evidence level, and what is unknown. The claim: train explosive power and you might live longer. The evidence: one large observational study. Observational means association, not proof of cause. Maybe powerful people are just healthier in ways the researchers could not fully adjust for.

Jordan: Exactly. What is unknown: whether power training itself changes your fate, how much power you need, and whether power beats cardio or just complements it. The researchers said this does not yet establish whether power outperforms cardiovascular fitness.

Alex: Now, the fun part. The biohacking world has been shouting about this for years. Dave Asprey, Doctor Mark Hyman, Gary Brecka, all of them have been telling people to lift heavy and move fast, to train strength and power as a longevity strategy. The mainstream is now catching up to what the alternative health crowd was saying a decade ago.

Jordan: Which does not mean everything in alternative health is right. It means sometimes the outsiders spot something real before the journals do. That is why we say beliefs are working knowledge. You update when the evidence arrives, and you stay humble about what you do not know.

Alex: And staying humble was literally the theme at the Longevity Investors Conference in Gstaad last week. Investors, scientists, and founders gathered to ask one question: what is science, and what is hype?

Jordan: The co-founder, Doctor Tobias Reichmuth, opened by saying the move into human trials is the defining development of the year. Then he said, quote, so far, we have not seen human life extension happening, urging investors to back evidence over marketing.

Alex: I love that. In a room full of people with money on the line, the honest sentence wins. And Professor Steve Horvath, the epigenetic clock pioneer, gave investors a filter for any claim of reversed biological age. Three Cs. Control group, context, and clock.

Jordan: Control group, context, and clock. That is a bumper sticker for critical thinking. If someone tells you a peptide or a protocol reversed their biological age by ten years, ask: compared to what control group, in what context, and measured by which clock? If they cannot answer, you have your answer.

Alex: That is the whole philosophy of this show in one story. Question what knowledge even means. Present the claim, show the evidence, admit the gaps, and update your beliefs when better data shows up. Whether it is AI agents or muscle power, the method is the same.

Jordan: Beautifully said. That is our bonus episode. DevDay's always-on dots, the White House self-policing pact, thirty billion dollars chasing OpenAI, and a reminder to move fast, literally.

Alex: We will be back Monday with the full weekly episode. Until then, keep questioning everything, including us.

Jordan: Especially us. See you Monday.
Question everything — including us… especially us.