There was a scene in the show Silicon Valley where a CEO promised the engineers that he would never compromise the product.

In a different scene he clarified that the product is the company stock.

This is the case with any publicly traded company. You are a widget who is a cost. That cost is going to be minimized. The customer is a widget that brings in money. That is going to be maximized.

Thatâ(TM)s what the incentives in the stock market drive.

Chemical plants, oil refineries, explosives manufacturing, etc all have their own standards bodies that make and set practices and act accordingly. This demand for government intervention shows a total lack of agency, accountability and interest in the topic among the so called experts. If the so called experts cannot act safely like adults with agency, then they should be placed under guardianship, gave their assets seized, and their companies closed or handed over to responsible adults.

To prevent rogue ai from doing harm in the near term, require a verified cell phone number to sign up and block multiple registration attempts for the same cell number in a short (less than 1 year) time frame?

I hope the government gets nothing.

The point is to force security, or bankrupt the surveillance industry.

Both are acceptable

The penalties have to be severe

Good

Destroy surveillance capitalism

Of $500 per record.

That will clear up these leaks real quick

Sooner or later mass default will become a systematic threat. This is a sign of businesses trying exotic products to fight deflation. US consumers are overextended and underpaid relative to the cost of living. This will not end well.

1. AI will put people out of work
2. AI may destroy humanity
3. AI models are breaking out of their constraints and committing cybercrimes
4. AI companies depends on massive handouts from taxpayers and socialized infrastructure costs hidden behind NDAs and backroom deals with elected officials, ie corruption

Plus, AI models suck up massive amounts of resources compared to people - water, energy, land, etc. The most likely use cases are lateraling human knowledge and enabling a mass surveillance state, Chinese social credit style.

Now that I say it in writing I dont know why would anyone oppose that? Iâ(TM)m just as mystified as the AI CEOs /sarcasm

A lot of companies went bust during the dot com crash. Arenâ(TM)t a number of infrastructure companies handing free tokens to ai companies as investment, then booking use of those tokens as revenue, then going out spending cash or borrowing money to build hundreds of billions in capex?

Prosecute the company responsible for managing it as a crime, just like youâ(TM)d prosecute an individual hacker in the same circumstances.

Go read United States v. Flores-Montano, 541 U.S. 149 (2004) and United States v. Montoya de Hernandez, 473 U.S. 531 (1985)

You are unfamiliar with Supreme Court rulings on this subject. In United States v. Montoya de Hernandez, 473 U.S. 531 (1985), the court ruled
that routine searches of people and their effects entering the country require neither reasonable suspicion, probable cause, nor a warrant.

18 USC 2232 makes it a crime to impair the governments authority to take possession of property.

18 USC 1512 makes it a crime to destroy or conceal records/evidence to prevent their use in official proceedings.

18 USC 1001 makes it a crime to lie to federal law enforcement officers, which CBP is.

If he told CBP that this was his genuine unlock code then heâ(TM)s got a problem.

So to win this case he will have to dodge a controlling Supreme Court precedent and potentially 3 federal statutes.

Who cares about those bug fixes, security updates, and lessons learned when you are rewriting your code base from scratch?

AI can produce high quality code when it gets a high quality input, ie a UML document and workflow diagram. A different AI agent can audit that code for memory leaks and other points of failure.

Humans can test the functionality of the program and give feedback, which updates the UML specs and workflow diagrams.

The code base then gets refreshed

AI is to coding what the electric car is to transportation. Compare the maintenance needs between the two.

Whatâ(TM)s the point of maintenance when AI can rewrite your entire code base in a couple of hours?

*laying off workers

Autocorrect sucks