We're in the Singularity Now — Here's What That Actually Means

Reporting on the Moonshots podcast conversation with Peter Diamandis, Salim Ismail, Alex Wiesner-Gross, and Dave Blundin.

---

The Pace Is No Longer Incremental. It's Exponential.

I sat with Peter Diamandis, Salim Ismail, Alex Wiesner-Gross, and Dave Blundin recently on the Moonshots podcast, and the conversation hit a point that clarified something I've been trying to articulate for months.

We are not waiting for the singularity. We are in it. Right now.

The evidence is not subtle. In the last 8 weeks alone, there have been 15 major AI model releases. Not iterations. Not tweaks. Major, capability-shifting releases from OpenAI, Google, Anthropic, DeepSeek, and others. Each one leapfrogging the last in ways that would have seemed impossible just months ago.

"Things are moving so quickly now," one of the hosts noted, "that on a month-by-month basis, we're able to see the hardest benchmarks creep up 1% per month."

That's not progress. That's acceleration. And it's compounding.

---

The Real Bottleneck: It's Not Software Anymore. It's Physics.

Here's what surprised me most in the conversation: the bottleneck to AI progress has shifted.

It's no longer about who can train the best model. It's about who can secure the most compute and the energy to power it.

Google just committed $40 billion to Anthropic. Amazon committed $33 billion. These aren't venture investments. These are infrastructure wars.

And the real constraint? "It's all bottlenecked at TSMC. That's the actual bottleneck to all of AI."

Think about that. The speed of AI progress is now determined not by how smart your engineers are, but by whether you can get semiconductor manufacturing capacity from a single company in Taiwan. This is a geopolitical chokepoint that will reshape the next decade.

The implication is stark: compute and energy are now the limiting factors. Model development capability has gotten so good that every major lab can build something competitive. But not every lab can secure the chips and power to run it.

---

Medicine Is Being Remade in Real-Time

One of the most concrete signals of how fast this is moving came from the medical AI discussion.

A new personalized cancer vaccine is showing 87.5% survival rates in early trials. That's not incremental. That's potentially transformative.

AI diagnostic systems are now outperforming human doctors. One system scored 59 versus 43.7 for human clinicians on a diagnostic benchmark. These aren't edge cases or cherry-picked examples. These are becoming routine.

And the shift isn't toward "AI assists doctor." The shift is toward "AI replaces doctor in specific domains."

Dave Blundin put it directly: "Of course, this is about replacing doctors. I mean, let's call a spade a spade."

This is happening now. Not in 10 years. Now.

---

The Economy Is Reorganising Around Token Costs

The conversation shifted to economics, and this is where it got genuinely strange.

Companies are now competing not on headcount but on "token spend" — how much they're spending on AI compute relative to human salaries. This has become a status symbol in startups. "Token maxing" is bragging that you're spending more money on AI compute than it would cost to hire human workers.

The implication: human labour, at least for white-collar work, is about to become a second-order concern in business models.

Peter Diamandis was blunt about the career implications: "Getting a job is the old model. The old model of do well in high school, get a good college, get a diploma, get hired as a junior person, and work your way up the chain. That is vaporized."

This isn't speculation. This is already being priced into company structures.

---

Government Operations Are Becoming Agentic

One of the most forward-looking examples came from the UAE's deployment of agentic AI across 50% of government services.

Not assisting civil servants. Replacing the need for human civil servants in specific workflows. The UAE is running a live experiment in what happens when you let AI handle routine government operations — licensing, permitting, regulatory processing.

The reason it works: "In prescriptive workflows you can absolutely completely automate."

Most government work is prescriptive. It follows rules. It's exactly what AI excels at.

The question isn't whether this is coming to democracies. It's how fast and whether citizens are ready for it.

---

Surveillance and Deepfakes: The Dark Side

Not everything in the conversation was optimistic.

New surveillance tools are emerging that make previous privacy breaches look quaint. OpenAI's Chronicle runs background agents taking periodic snapshots of everything on your screen. These systems are getting more invasive, and the regulatory frameworks are nowhere near.

Deepfake technology is advancing faster than deepfake detection. "Deep fake verification may no longer be reliable," one host noted. Which means you can no longer assume you can tell the difference between real and synthetic media.

This creates a strange dynamic: we have the technology to solve surveillance and verification problems, but implementation lags behind need. And the incentives are misaligned — nobody profits from solving this faster than the alternative.

---

Reframing the Singularity: It's Now, and It's an Interval

This is where the conversation crystallised something important.

Ray Kurzweil predicted the singularity would happen in 2045. Most people imagine it as a point in time — the moment when AI becomes superintelligent and everything changes.

But that's not what's happening. Alex Wiesner-Gross captured it perfectly: "My version of the singularity isn't a point in time, and it's certainly not in 2045. It's now and it's an interval and we're right in the middle of it."

We're not approaching an inflection point. We're living through one. The singularity isn't an event. It's a process. And we're in it.

This reframes everything. It means:

• The career planning advice you got 5 years ago is already obsolete
• Medical care is being remade in real-time
• Government operations are being automated faster than policy can adapt
• Economic structures are shifting beneath everyone's feet
• Geopolitical power is concentrating around compute access and energy

---

What This Means for You

If you're reading this and feeling vertigo, that's appropriate. The pace of change is legitimately unprecedented.

But here's what I took from the conversation:

First: The window for understanding how AI works is closing. In 6 months, this will be even more complex and move faster. If you want to build intuition about what's happening, now is the time.

Second: Career planning requires a complete rethink. The old model (school → college → entry-level job → climb the ladder) is broken. You need to think about developing skills that complement AI, not compete with it. And you need to think about building your own leverage faster than waiting for a title.

Third: The geopolitical struggle over compute will reshape everything. TSMC, energy infrastructure, semiconductor access — these are the new battlegrounds. Whoever controls the ability to run AI at scale controls the next decade.

Fourth: Medical breakthroughs are happening in real-time. If you care about longevity, this is the moment to be paying attention. The tools are being built now. The protocols are being discovered now.

Fifth: We are living through the most transformative period in human history. Not approaching it. Living through it. That's not hype. That's pattern-matching against what the data is showing us.

---

Listen to the full conversation on the Moonshots podcast with Peter Diamandis, Salim Ismail, Alex Wiesner-Gross, and Dave Blundin.

Lisa Tamati reports on AI, technology, and the future of human potential at PTLsignal.com. This analysis is for informational purposes only and does not constitute investment advice.