“AI Allows Anybody to Do a Job.” One Word in That Sentence Is Wrong. And It Changes Everything.

Shawn Kanungo posted something a few weeks ago that’s still circulating: “AI allows anybody to do a job.” Shawn Kanungo posted something a few weeks ago that’s still circulating: ‘AI allows anybody to do a job.’ That claim raises an AI judgment question most people are skipping past.

Shawn is a keynote speaker I’ve featured myself, and he’s not wrong about the capability. OpenAI’s newest model, GPT-6 Astra, can run professional software it was never specifically trained on. Shawn tested it himself, on vacation: raw multi-camera, multi-track footage, one instruction, sync it, edit it in Final Cut Pro, export ten optimized vertical clips, caption all of them. It did it, unaided, like a professional. He’s also run it through Adobe After Effects, Blender, and Unreal Engine, the same combination OpenAI used in its own launch demos. That’s real, and it’s a genuine step change, on OpenAI’s own computer-use benchmark, Astra scored 72.6% in about 40 minutes per task, versus 65.7% at roughly 75 minutes for the model it replaced.

His conclusion, AGI is here. It’s worth being precise about that word, because OpenAI’s own announcement never uses it. The company calls Astra “a new generation of intelligence.” Whether that gap between the label and the launch matters is a separate conversation. What matters more is the sentence Shawn actually built his conclusion on.

The AI Judgment Question Hiding in One Word

AI doesn’t allow anybody to do a job. It allows anybody to operate the software behind a job. Those are not the same claim, and the difference is the whole point.

Operating Final Cut Pro was never the job. The job was knowing what the footage actually told, what to cut, and why. That judgment is what’s still standing once software fluency stops being scarce, and it’s the one thing this model, or the next one, still hasn’t done for anyone.

The Split Reaction: Both Sides Missed the AI Judgment Question

Within about a day of Astra’s launch, a second, opposite reaction was circulating on the same release, a post dismissing the AI-optimism crowd as “pitchmen,” linking a satire video about AI and losing what makes us human. Two credible, engaged reactions to the identical release. One celebratory, one dystopian, both confident.

Neither reaction gives you a Monday morning plan. That’s not a criticism of either camp’s sincerity, it’s a description of what’s missing from both. The question was never whether you should be excited or scared that a model can now run software it has never seen before. The question is whether you’ve built the one thing it still can’t do for you.

Why This Is a TALENT Problem, Not Just a Capability Story

The Kryptonite Defense names TALENT as one of five ingredients an organization needs to withstand the 7-Sided Pincer Movement. When a model can operate professional software at a professional level without being specifically trained on it, software fluency stops functioning as a credential. Judgment, knowing what’s worth making and why, is the ingredient still standing once execution gets cheap.

It’s also an IDEAS story. Once execution is nearly free, creative direction and intent become the actual bottleneck, not the tool. And it’s a DISTINCTION story in a quieter way: professional tool fluency stops differentiating a person or an organization from anyone else running the same model.

The 5-Ingredient Kryptonite Defense: Choosing a Side Deliberately

Neither declaring AGI has arrived nor mourning what AI is taking from us is a decision. Both are reactions. Those prepared need not fear the forces at work. Here is how each of the five ingredients builds the judgment that actually decides which side of this you’re on.

IDEAS creates competitive advantage because once a model can execute almost anything, original thinking about what’s worth executing becomes the scarce resource, not the execution itself.

SPEED creates competitive advantage by closing the gap between a capability like this becoming available and your organization actually building the judgment to use it well, rather than reacting to the headline for a news cycle and moving on.

TALENT creates competitive advantage because knowing what footage tells the real story, or what result is actually worth producing, is a human judgment call a model still can’t make for you, no matter how well it operates the software.

DISTINCTION creates competitive advantage by making the outcome of your judgment, not your tool fluency, the thing that sets your work apart from anyone else with access to the same model.

LEADERSHIP AT ALL LEVELS creates competitive advantage because the choice between the AGI-declaration reaction and the panic reaction gets made individually, at every level of an organization, every time someone sees a release like this one. Judgment has to be a habit built at all of those levels, not a talking point from the top.

The Question to Take Back to Your Organization

A model can now edit your video, animate your scene, and score your soundtrack, unaided. What remains open is whether your organization has built the AI judgment to decide what’s worth making with it, and why. That much is no longer a question. The open question is whether your organization has built the judgment to decide what’s worth making with that capability, and why, or whether it’s still deciding, reaction by reaction, which camp it’s in.

Curious where your organization actually stands on this? The Kryptonite Scorecard is an outside measurement of where you stand, not a number you assign yourself, about 15 minutes.

And if this resonated, Distinct or Extinct lays out the full framework these ingredients come from.

 

Related Reading

“OpenAI’s Newest Model Knows When It’s Being Tested. So Did Volkswagen’s Diesels.” 

“OpenAI’s Own Chief Scientist Just Said Nobody Is Prepared for What’s Coming.”