Pucker time ...

Unit Circle

 The specter of unfettered AI becoming reality becomes possible because the notion of ethics was never addressed when the Googles of the world created tech that will forever remain unknown as code has to write code in real time in order for the tech to react to the real world in real time. 

Actual meaning was also never factored into the creation of AI in addition to ethics as understanding how to integrate this complex dataset to open ended tech in reasonable time frames was considered too hard to do or better yet, when the time comes, Google and company can just plug it in as needs warrant because the AI in question will be mature enough to make it happen but this hoped for eventuality really doesn't make sense, does it? 

Besides, pattern recognition to the nth degree works better without all that excess baggage, right?

Pucker time ...

The tech industry now finally realizes the horse isn't just out of the barn; the barn has been burned to the ground to power the data center. Eating the world applies and ... the panic is palpable.

Tameem Sankari for Noema Magazine

A recent piece in Noema Magazine highlights a new, proposed solution to the AI alignment problem: crowdsource a global moral framework.

Using these companies’ best models to ask everyday people urgent questions about AI — What do you want this technology to do for you, if anything? What should it never be allowed to do? Who should govern this technology and according to what principles? — is not a thought experiment about hypothetical capacity. The infrastructure to ask hundreds of millions of people or more the same question in the same week
already exists.

“Whether the danger is an AI slipping out of human hands or sovereignty slipping out of governments’ grasp, the fix is the same: More people, not fewer, need a say before consequential decisions are made.”

 Together, the best AI models out there should be able to produce something like a global constitution for AI that most people in the world would find acceptable — an idea that Michiel Bakker, a co-author of“Europe 2031,” first raised in a working session at the Superintelligence Summit at the Council on Foreign Relations
in March.

The idea, in this piece, is the possibility we can use AI to act as a super-mediator, synthesizing the opinions of hundreds of millions of people to magically produce a "global constitution for AI."

It all sounds doable but if we pull back the curtain and look at how AI works, we have to ask hard questions.

1. Does consensus equal truth?

If an AI is programmed to find the "common ground" among a hundred million people, does it actually establish a moral baseline, or does it just mathematically calculate the path of least resistance? When you synthesize that much data, you don't necessarily get a moral constitution; you get a highly optimized statistical average of human bias and anxiety. Truth and structural reality, unlike politics, are rarely a popularity contest.

2. Who controls the scoreboard?

Even an AI super-mediator needs a prompt. It needs a goal to optimize for. Who defines what a "good" or "acceptable" consensus looks like? If the algorithmic goal is simply to "minimize conflict," the AI will naturally suppress the high-friction, lateral human intuition that actually drives breakthroughs. It enforces a sterile, frictionless compliance.

3. Can you bolt on a soul?

A constitution is a semantic, deeply human document. It relies on the weight of physical reality and shared history.
Can a machine that doesn't experience physical reality and pays no real cost for its own decisions, truly mediate
human morality?

But a glimmer ...

The tech giants are trying to crowdsource the ethics they forgot to build into the foundation. But as any engineer will tell you, bolting ethics onto a system after it has already been deployed is like putting a 5th wheel on a car.
It's useless. Form follows function.

Trust and ethics cannot be retroactively calculated by a large language model. They have to be hardwired into the system on Day One. This is a wicked problem. The minute you try to solve it, the problem changes. Can a semantic framework be applied?  Can ethics be actually integrated as trying to apply it as an afterthought is a non starter. What about an organic guide providing a steady, common-sense anchor of not being a jerk protocol. Is any of this possible?

4. One never knows.

The infrastructure for safe AI isn't a global poll. It is a one-to-one semantic bridge based on meaning, not just pattern recognition. Some people are finally waking up, frightened when viewing this reality but they are looking for the solution in the exact same brute-force technology that caused the problem in the first place.

Pucker time applies but additionally, time's running out. The paperclip scenario looms.

Waining Gibbous