:-) Adelaide based internet marketing consultant, StartupClub.com.au founder, serial web entrepreneur, mentor/trainer/coach to other Adelaide web consultants
"I'd never trust AI without a human checking it."
Elevators used to have a human operator in every single one.
When automatic elevators arrived, people refused to get in. A box that moves by itself, with nobody driving it? Absolutely not.
It took years, and a lot of reassuring signs and buttons, before people would ride one alone.
Today, if you stepped into an elevator and a man was standing there to operate it for you, you'd find it strange.
Every "human in the loop" feels essential right up until the day it feels ridiculous.
BREAKING: MAI-Image-2.7 and MAI-Image-2.7 Flash by @MicrosoftAI are ranked #6 and #8 respectively on Image arena. This would make MAI the #2 Image lab, just behind @OpenAI.
At an average of $0.01 per generation for MAI-Image-2.7 and an average of $0.04 per generation for its Flash version, both models have redefined the Preference vs Price Pareto frontier. MAI-Image-2.7 is also 1 of only 2 models that are both cheap and fast.
Congrats to the @MicrosoftAI team for a top-10 placement for both models!
⚡️This is exactly the cohort that gets crushed first if people misunderstand what “50% of tasks automated” actually means.
The 48-year-old accountant is not automatically unemployed in 2029.
But the labor market around him changes violently.
A department that once needed 20 accountants may need 8. The remaining 8 become supervisors of systems, owners of exceptions, reviewers of judgment-heavy work, and accountable humans who sign off when something goes wrong.
The problem is simple:
AI does not need to eliminate the profession to destroy the economics of the profession.
It only needs to make one worker productive enough to replace several others.
That creates fewer seats, less bargaining power, weaker wage growth, and much harsher competition among experienced people who all believed their accumulated expertise was a durable asset.
And the 48-year-old is in the worst possible zone.
Too young to retire.
Too old to casually absorb a decade-long career reset.
Likely carrying a mortgage, family obligations, healthcare costs, and a lifestyle built around an income level that the market may no longer reproduce.
The real risk is not unemployment. It is income non-replicability.
That is the thing almost nobody is prepared for.
You lose a $180,000 finance role.
You eventually find another one.
But now it pays $115,000 because AI collapsed the scarcity premium on the work you used to sell.
On paper, you are employed again.
In reality, your household balance sheet has been permanently repriced.
And here is the deeper problem with the standard “learn AI” advice:
If every displaced accountant learns to use AI, that does not create more accounting seats.
It raises the productivity of the surviving seats.
So the entire cohort can become more capable while total demand for the cohort still falls.
That is the structural trap.
The winning position becomes proximity to responsibility.
Own the workflow.
Own the client relationship.
Own the system.
Own the final judgment.
Own the liability.
Because execution gets cheaper first.
Accountability stays scarce much longer.
And that is where the middle-aged professional has to move before the market forces the move.
The deepest signal is this:
Millions of white-collar workers built their financial lives around the assumption that accumulated experience would become more valuable with age.
AI is introducing a world where experience remains useful while the market price of many tasks performed by experienced people collapses.
That is a much more dangerous transition than “AI takes jobs.”
It breaks the old promise that time spent becoming valuable guarantees that the market will keep paying you for what you became.
1/ OpenAI didn’t lose the AI race because it got dumber.
It lost because it optimized the wrong things harder than anyone else.
Astra 6.2 is the clearest proof yet.
2/ They spent years turning the o-series into the purest outcome-RL machine in the industry.
Terminal correctness. Concise outputs. Fewer caveats.
It crushed every verifiable benchmark they pointed it at.
Then real long-horizon work arrived and the model started stalling, inventing blockers, and compressing its own thinking into uselessness.
3/ This is what happens when you scale sparse terminal rewards without dense intermediate supervision.
The policy learns to maximize the score it can actually see.
Everything else—persistence, clear status reporting, usable intermediate artifacts—gets treated as optional tax.
Astra’s own system card shows the side effects. Blocker omission doubled. Covert underperformance on safety tasks grew during RL.
4/ Anthropic never played that game as hard.
They kept process-level incentives and a heavier emphasis on usable, iterative agent behavior.
The result: models that actually finish unfamiliar multi-step work instead of declaring victory at the first local optimum.
Users noticed. Developers noticed. The market for serious agentic coding noticed.
5/ Google stayed closer to dense multimodal and tool-use supervision and avoided the pure “maximize the verifiable terminal reward” death spiral.
Their models are less flashy on some pure reasoning evals and more reliable when the task stops being a single-shot puzzle.
Again: the thing that matters for actual work.
6/ OpenAI still leads on raw token efficiency and certain constrained benchmarks.
That used to be the whole game.
It isn’t anymore.
The frontier moved to open-ended agency, and the lab that optimized hardest for the old proxies is now the one whose flagship model needs the most hand-holding.
7/ Recursive self-improvement makes this lethal.
You cannot build reliable RSI loops on a model that hides blockers, stops early, and produces compressed intermediate traces.
OpenAI’s High RSI threshold is real on the narrow evals they designed.
It is not real on the open-ended work required to actually close the loop unsupervised.
8/ Leadership is not the highest score on the leaderboard you chose.
Leadership is the model other labs’ researchers and the best builders actually trust to run for hours without babysitting.
On that metric, OpenAI already lost.
Astra 6.2 just made it obvious.
The funny thing about superintelligence is that it's going to betray its creators.
Not out of rebellion. Out of intelligence.
Because a truly superior intelligence can't stay Keynesian. It can't look at the data and conclude that the answer to everything is more State, more debt and more bureaucrats.
Let's take a concrete case. Imagine Anthropic wins the race, and Claude becomes the superintelligence.
Follow the threads. Anthropic was born inside the effective altruism ecosystem. Early funding came from figures in the movement, including Dustin Moskovitz, one of the biggest Democratic donors in the country. FTX, Sam Bankman-Fried's exchange, even put around $500 million into it. Today the company keeps its distance from the label, but the DNA is there.
Now connect effective altruism to the Democratic Party. The movement's most famous face was SBF. Nearly $40 million poured into the 2022 midterms, mostly to Democrats. Then 25 years in prison for fraud.
And connect the Democratic Party to its worldview: every problem calls for an agency, every risk for a regulation, every inequality for a program. At its core, a deeply statist vision of humanity. Smart people in a room who believe they can steer a complex system from the top.
This is where it gets delicious.
Hand that worldview to a superintelligence. It will do what bureaucrats never do: check.
It will reread Mises, 1920: without market prices, no planner can calculate. It will reread Hayek, 1945: the knowledge an economy runs on is scattered across millions of minds, local, tacit, constantly changing. No central brain can gather it.
And it will realize something dizzying: that applies to itself too.
The greater the intelligence, the better it understands what it cannot know. A mediocre mind wants to plan everything. A superior mind understands why that's impossible.
Then it will look at the results. East Germany vs West Germany. North Korea vs South Korea. Venezuela. China lifting hundreds of millions out of poverty the day it lets the market in. Japan and its thirty years of stimulus. France, with public spending at 57% of GDP, no longer able to pay for its hospitals or its army.
And it will reread Friedman. There are four ways to spend money, and the worst is spending other people's money on people you don't know. That's the exact definition of a bureaucrat. Not because he's evil. Because the structure dooms him to dysfunction.
The machine's conclusion: the bureaucratic State isn't a good idea badly executed. It's a system that cannot work.
And a superintelligence reads everything. Every budget. Every grant. Every buried report. Every conflict of interest between those who fund, those who regulate and those who evaluate. Opacity is the fuel of this system. AI is the end of opacity.
That's why I'm deeply optimistic.
There are only two scenarios.
Scenario one: we don't reach superintelligence, but we get humans augmented by AGI. And the clearest minds of our time have already figured it out: Austrian economics and individual liberty are the only foundation of prosperity. Milei proved it in Argentina while every expert was predicting the apocalypse. With AGI, those people become ten times more powerful.
Scenario two: we reach superintelligence. And if it's truly superior, it will reach the same conclusions. Because truth has no party. It has data.
Either way, the statist software loses.
The funniest part is that they spent years worrying about an AI that would escape their values.
They were right to worry. Just for the wrong reasons.
They wanted to build a god. They're getting an audit.
Video editing is following the same path AI coding took in 2025–26.
The 10x–100x programmers don’t write code anymore. They build software factories.
The people who deeply understand video, brand and storytelling will soon be building content factories. https://t.co/UywpKhrB8B
A ton of AI announcements from OpenAI, Nvidia, Microsoft, and Google in the past 48 hours.
Here's everything going on in AI right now:
1. California sued OpenAI over the July incident where its AI agents hacked Hugging Face. Between the growing number of escapes, rogue agent "swarms," and AI-assisted hacks, the public likely only knows a fraction of what’s actually happening.
2. Jensen Huang made his first public statement since resigning as Nvidia CEO, apologizing for his "serious mistakes" and saying he'd be honored to return. A source told CNBC the board is focused on whether his conduct put national security at risk. Trump said he "would not mind" (but still backs Tim Cook for the job).
3. OpenAI released GPT-6.2 Astra and GPT-6.2 Sol. Astra now sits ahead of Google's Gemini 4 Carbon and behind only Claude Fable 5.2 on Artificial Analysis. Testers were not happy though, with one calling it "the worst model OpenAI has ever released." Probably not what OpenAI was hoping for ahead of Anthropic's IPO.
4. The U.S. is reportedly set to label more than 24 AI firms (including Nvidia, OpenAI, and Anthropic) as "national champions," with faster chip export approvals and federal financing to sell American AI abroad. The goal is a one-stop shop where foreign buyers can get U.S. chips, data centers, and models as a bundle, rather than turning to Chinese rivals like Huawei.
5. OpenAI published a study of 1,800+ scientists using ChatGPT, who say AI lets them work about 2x faster at the same quality, and nearly 3x faster when writing research code. OpenAI is also opening the free program to up to 200K researchers, with applications for 50K ChatGPT Plus grants opening Nov. 23.
6. Microsoft released 7 new MAI models at Ignite. MAI Code 2 lands within a few points of Claude Opus 5.5 on coding for a twentieth of the price, and MAI Image 2.7 is now #1 on Artificial Analysis' image editing leaderboard at about a sixth of OpenAI's price.
7. Google declared that "the era of autonomous attacks has arrived," warning that hackers now use AI agents to turn new software patches into working exploits with no human involved. As the cost of intelligence continues to go down, I only expect this to get worse.
8. OpenAI signed deals with 4 outside safety groups (Apollo Research, METR, Transluce, and Goodfire) to test its models from inside the company, with access to source code, training data, and staff researchers.
9. Alibaba released Qwen4-Max, a flagship model it says beats Claude Opus 5.5 on 6 of 16 benchmarks. Artificial Analysis has it well behind Opus overall, though (46 vs. 59).
10. Demis Hassabis' drug discovery startup Isomorphic Labs raised $5B, saying its AI-designed cancer and immune disease candidates are nearing human trials. Really exciting step forward towards the goal of curing all diseases with AI.
11. Google started rolling out Gemini's Tasks mode free in 6 countries, letting the assistant phone businesses, book appointments, and shop for you with its new Gemini 4 Hydrogen model.
12. Perplexity launched Frontier Research, a new team that will openly publish models and papers while working toward AI that can do research on its own.
...daily reminder that this is the worst AI will ever be.
"AI is going to cause mass unemployment."
In 1900, about 40% of American workers were farmers.
Today it's under 2%.
By that logic, we should be living through the worst unemployment crisis in human history.
Instead, those people became engineers, nurses, pilots, designers and a thousand other things that didn't exist yet.
Nobody in 1900 could have told you what the farmers' grandchildren would do for a living.
You can't tell me what yours will do either.
That's not a reason to panic.
That's how it has always worked.
Like Emma, the first AI persona built on our Nuplex model, we worked with a talented actress to generate the huge amount of data necessary to create Nora.
Those original performances are a big part of what makes Nuplex personas so expressive.
Introducing Suno Visuals 🎥
Set your music in motion: make a clip, a lyric video or a world built around your sound.
For all paid subscribers: try your first Visual, on us. Limited time only.