It’s getting hard to keep up with all the reports filed about AI companies that operate as predatory criminal enterprises desperate to maximize rapid growth at others’ expense.

VentureBeat

…LLM companies such as OpenAI, Anthropic, Cohere and even Meta — traditionally the most open source-focused of the Big Tech companies, but which declined to release the details of how LLaMA 2 was trained — have become less transparent and more secretive about what datasets are used to train their models. […] …there is no longer any doubt that copyright infringement is rampant. As companies seeking commercial success get ever-hungrier for data to feed their models, there may be ongoing temptation to grab all the data they can.

Shout out to Anthropic as indistinguishable from OpenAI, after it was founded by OpenAI staff to be distinguishable.

Vox

…the bloom is coming off the AI-generated rose. Governments are ramping up efforts to regulate the technology, creators are suing over alleged intellectual property and copyright violations, people are balking at the privacy invasions (both real and perceived) that these products enable, and there are plenty of reasons to question how accurate AI-powered chatbots really are and how much people should depend on them. Assuming, that is, they’re still using them. Recent reports suggest that consumers are starting to lose interest.

Washington Post

Behind the AI boom, an army of overseas workers in ‘digital sweatshops’ … In the Philippines, one of the world’s biggest destinations for outsourced digital work, former employees say that at least 10,000 of these workers do this labor on a platform called Remotasks, which is owned by the $7 billion San Francisco start-up Scale AI. Scale AI has paid workers at extremely low rates, routinely delayed or withheld payments and provided few channels for workers to seek recourse, according to interviews with workers, internal company messages and payment records, and financial statements. Rights groups and labor researchers say Scale AI is among a number of American AI companies that have not abided by basic labor standards for their workers abroad.

New York Magazine in collaboration with The Verge

This tangled supply chain is deliberately hard to map. According to people in the industry, the companies buying the data demand strict confidentiality. (This is the reason Scale cited to explain why Remotasks has a different name.) Annotation reveals too much about the systems being developed, and the huge number of workers required makes leaks difficult to prevent. Annotators are warned repeatedly not to tell anyone about their jobs, not even their friends and co-workers, but corporate aliases, project code names, and, crucially, the extreme division of labor ensure they don’t have enough information about them to talk even if they wanted to. (Most workers requested pseudonyms for fear of being booted from the platforms.) Consequently, there are no granular estimates of the number of people who work in annotation, but it is a lot, and it is growing. A recent Google Research paper gave an order-of-magnitude figure of “millions” with the potential to become “billions.”

Pay extremely low rates, routinely delay or withhold payments, and illegally redirect wealth from everyone to a few.

Digital dictatorships.

A few hours ago the SF Chronicle published a map of crashes that illustrates quite well a failure of driverless cars to deliver safety or reliability.

There are so many simple yet catastrophic failures it’s hard to choose which one will become most popular among the many groups watching and aiming to disrupt transit in major urban areas.

For example an easily predictable congestion of wireless communications led to fleets of cars going into failure mode, stopping and blocking all traffic as if robots staging a protest. SFPD had to be diverted from real work to attend to giant incapacitated and needy robots, ultimately redirecting traffic to other streets that weren’t in crisis.

As many as a dozen stalled Cruise autonomous vehicles blocked streets in San Francisco’s North Beach and near this weekend’s Outside Lands music festival, snarling traffic and frustrating riders barely a day after state regulators voted to allow the unlimited expansion of robotaxi companies.

Social media users posted about one incident late Friday in which about 10 Cruise vehicles appeared to be standing still with their hazard lights flashing, blocking lanes on Vallejo Street near Grant Avenue.

The whole city is vulnerable to sudden remotely controlled shutdowns. But more to the point, the map of crashes shows the robots are failing at basic daily safety before we even get to the phase of trivial targeted wave attacks on them (e.g source code).

In 2019 we all watched the sleazy car salesman tactics get 250,000 people to pay $100 for nothing.

Worse than nothing, they paid for the promise of a “tough” truck that immediately was revealed as fragile.

You may remember LEGO cleverly mocked this slime spectacle with what seemed to be a far superior toy truck design.

To put it another way, context matters here. LEGO puts a huge amount of engineering and careful craftsmanship into their vehicle replicas. Their recreations of famous cars are truly impressive at any scale.

So when LEGO threw together a minimal effort block they described as an improved version of the silly Tesla Truck design craze, it was literal mockery of inflated egos at Tesla peddling sadly simplistic ideas and low skills. LEGO slam dunked on the spectacle, wisely foreshadowing the truck’s predictable failures.

FastCompany is now laughing out loud at the little dictator running Tesla, after he just threw up his hands and issued an edict that the Truck must be built like a LEGO.

The problem, according to Musk, is the bright metal construction and predominantly straight edges mean that even minor inconsistencies become glaringly obvious. To avoid this, he commanded unparalleled precision in the manufacturing process, stating in his email that “all parts for this vehicle, whether internal or from suppliers, need to be designed and built to sub 10 micron accuracy. That means all part dimensions need to be to the third decimal place in millimeters and tolerances need [to] be specified in single digit microns.” …Musk added, “If LEGO and soda cans, which are very low cost, can do this, so can we.”

Commanded? Demanded? Unhinged.

If LEGO and soda cans can do this, why can’t a flamethrower at 100 meters perfectly turn an apple on my head into a delicious pie? I command you peons to make my fantasy a reality and if you fail I’ll just find more peons who keep believing.

Herr Musk seems raised on the privilege of an unrelenting pursuit of selfish fantasy and unable to grasp basic reality. His toddler-like curations of design based on mysticism, as if they could replace actual engineering knowledge, soon may have his legions of unskilled enablers/believers headed for a rough and abrupt awakening.

What do you call it when a giant flat shiny steel panel after three years still produces the exact opposite effect of what was promised to a quarter-million people who put money down?

Advance fee fraud truck.

The dumb design promised to be on the road by 2021 is a failure by almost every measure, a monument to a sheltered elitist South African apartheid boy pushing symbolism over substance. America should take down the 1920’s statues of General Lee and mount the 2020’s Cyber Truck on columns instead. Start renaming the overtly racist failure of Lee Street to Cyber Truck Lane. Same stuff, lessons not learned, 100 years later.

At this point you have to ask how a car company can exist let alone be valued when it so very obnoxiously shows it can’t handle even the basics of car design.

Studebaker folded for less.