Daily Archives: December 4, 2023

Brazillian city enacts an ordinance that was written by ChatGPT – might be first law entered by AI

Posted on by

City lawmakers in Brazil have enacted what appears to be the nation’s first legislation written entirely by artificial intelligence — even if they didn’t know it at the time.

The experimental ordinance was passed in October in the southern city of Porto Alegre and city councilman Ramiro Rosário revealed this week that it was written by a chatbot, sparking objections and raising questions about the role of artificial intelligence in public policy.

Rosário told The Associated Press that he asked OpenAI’s chatbot ChatGPT to craft a proposal to prevent the city from charging taxpayers to replace water consumption meters if they are stolen. He then presented it to his 35 peers on the council without making a single change or even letting them know about its unprecedented origin.

“If I had revealed it before, the proposal certainly wouldn’t even have been taken to a vote,” Rosário told the AP by phone on Thursday. The 36-member council approved it unanimously and the ordinance went into effect on Nov. 23.

“It would be unfair to the population to run the risk of the project not being approved simply because it was written by artificial intelligence,” he added.

[…]

“We want work that is ChatGPT generated to be watermarked,” he said, adding that the use of artificial intelligence to help draft new laws is inevitable. “I’m in favor of people using ChatGPT to write bills as long as it’s clear.”

There was no such transparency for Rosário’s proposal in Porto Alegre. Sossmeier said Rosário did not inform fellow council members that ChatGPT had written the proposal.

Keeping the proposal’s origin secret was intentional. Rosário told the AP his objective was not just to resolve a local issue, but also to spark a debate. He said he entered a 49-word prompt into ChatGPT and it returned the full draft proposal within seconds, including justifications.

[…]

And the council president, who initially decried the method, already appears to have been swayed.

“I changed my mind,” Sossmeier said. “I started to read more in depth and saw that, unfortunately or fortunately, this is going to be a trend.”

Source: Brazillian city enacts an ordinance that was written by ChatGPT | AP News

One AI image needs as much power as a smartphone charge

Posted on by

In a paper released on arXiv last week, a team of researchers from Hugging Face and Carnegie Mellon University calculated the amount of power AI systems use when asked to perform different tasks.

After asking AIs to perform 1,000 inferences for each task, the researchers found text-based AI tasks are more energy-efficient than jobs involving images.

Text generation consumed 0.042kWh while image generation required 1.35kWh. The boffins assert that charging a smartphone requires 0.012kWh – making image generation a very power-hungry application.

“The least efficient image generation model uses as much energy as 950 smartphone charges (11.49kWh), or nearly one charge per image generation,” the authors wrote, noting the “large variation between image generation models, depending on the size of image that they generate.”

The authors also measured carbon dioxide created by different AI workloads. As depicted in the graphic below, image creation topped that chart

screenshot_graph

Click to enlarge

You can read the full paper here [PDF].

Source: One AI image needs as much power as a smartphone charge • The Register

A(I) deal at any cost: Will the EU buckle to Big Tech?

Posted on by

Would you trust Elon Musk with your mortgage? Or Big Tech with your benefits?

Us neither.

That’s what’s at stake as the EU’s Artificial Intelligence Act reaches the final stage of negotiations. For all its big talk, it seems like the EU is buckling to Big Tech.

EU lawmakers have been tasked with developing the world’s first comprehensive law to regulate AI products. Now that AI systems are already being used in public life, lawmakers are rushing to catch up.

[…]

The principle of precaution urges us to exercise care and responsibility in the face of potential risks. It is crucial not only to foster innovation but also to prevent the unchecked expansion of AI from jeopardising justice and fundamental rights.

At the Left in the European Parliament, we called for this principle to be applied to the AI Act. Unfortunately, other political groups disagreed, prioritising the interests of Big Tech over those of the people. They settled on a three-tiered approach to risk whereby products are categorised into those that do not pose a significant risk, those that are high risk and those that are banned.

However, this approach contains a major loophole that risks undermining the entire legislation.

Like asking a tobacco company whether smoking is risky

When it was first proposed, the Commission outlined a list of ‘high-risk uses’ of AI, including AI systems used to select students, assess consumers’ creditworthiness, evaluate job-seekers, and determine who can access welfare benefits.

Using AI in these assessments has significant real-life consequences. It can mean the difference between being accepted or rejected to university, being able to take out a loan or even being able to access welfare to pay bills, rent or put food on the table.

Under the three-tiered approach, AI developers are allowed to decide themselves whether their product is high-risk. The self-assessment loophole means the developers themselves get to determine whether their systems are high risk akin to a tobacco company deciding cigarettes are safe for our health, or a fossil fuel company saying its fumes don’t harm the environment.

[…]

Experience shows us that when corporations have this kind of freedom, they prioritise their profits over the interests of people and the planet. If the development of AI is to be accountable and transparent, negotiators must eliminate provisions on self-assessment.

AI gives us the opportunity to change our lives for the better. But as long as we let big corporations make the rules, we will continue to replicate inequalities that are already ravaging our societies.

Source: A(I) deal at any cost: Will the EU buckle to Big Tech? – EURACTIV.com

OK, so this seems to be a little breathless – surely we can put in a mechanism for EU checking of risk level when notified of a potential breech, including harsh penalties for misclassifying an AI?

However, the discussions around the EU AI Act – which had the potential to be one of the first and best pieces of regulation on the planet – has now descended into farce since ChatGPT and some strange idea that the original act did not have any provisions for General Purpose / Foundational AI models (it did – they were high risk models). The silly induced discussions this has provoked has only served to delay the AI act coming into force for over a year – something that big businesses are very very happy to see.

23andMe hackers accessed DNA information on millions of customers using a feature that matches relatives

Posted on by

An SEC filing has revealed more details on a data breach affecting 23andMe users that was disclosed earlier this fall. The company says its investigation found hackers were able to access the accounts of roughly 0.1 percent of its userbase, or about 14,000 of its 14 million total customers, TechCrunch notes. On top of that, the attackers were able to exploit 23andMe’s opt-in DNA Relatives (DNAR) feature, which matches users with their genetic relatives, to access information about millions of other users. A 23andMe spokesperson told Engadget that hackers accessed the DNAR profiles of roughly 5.5 million customers this way, plus Family Tree profile information from 1.4 million DNA Relative participants.

DNAR Profiles contain sensitive details including self-reported information like display names and locations, as well as shared DNA percentages for DNA Relatives matches, family names, predicted relationships and ancestry reports. Family Tree profiles contain display names and relationship labels, plus other information that a user may choose to add, including birth year and location. When the breach was first revealed in October, the company said its investigation “found that no genetic testing results have been leaked.”

According to the new filing, the data “generally included ancestry information, and, for a subset of those accounts, health-related information based upon the user’s genetics.” All of this was obtained through a credential-stuffing attack, in which hackers used login information from other, previously compromised websites to access those users’ accounts on other sites. In doing this, the filing says, “the threat actor also accessed a significant number of files containing profile information about other users’ ancestry that such users chose to share when opting in to 23andMe’s DNA Relatives feature and posted certain information online.”

[…]

Source: 23andMe hackers accessed ancestry information on millions of customers using a feature that matches relatives

The disturbing part of this is that the people who were hacked were idiots anyway for re-using their password and probably didn’t realise that they were giving away DNA information about not only themselves, but their whole family to 23andMe, who sold it on. Genetic information is the most personal type of information you have. You can not change it. And if you give it to someone, you also give away your family. Now it wasn’t just given away, it was stolen too.

Electric Vehicles Are79%  Less Reliable Than Conventional Cars

Posted on by

Electric vehicle owners continue to report far more problems with their vehicles than owners of conventional cars or hybrids, according to Consumer Reports’ newly released annual car reliability survey. The survey reveals that, on average, EVs from the past three model years had 79 percent more problems than conventional cars. Based on owner responses on more than 330,000 vehicles, the survey covers 20 potential problem areas, including engine, transmission, electric motors, leaks, and infotainment systems.

“Most electric cars today are being manufactured by either legacy automakers that are new to EV technology, or by companies like Rivian that are new to making cars,” says Jake Fisher, senior director of auto testing at Consumer Reports. “It’s not surprising that they’re having growing pains and need some time to work out the bugs.” Fisher says some of the most common problems EV owners report are issues with electric drive motors, charging, and EV batteries.

Source: Electric Vehicles Are Less Reliable Than Conventional Cars – Consumer Reports

Plants may be absorbing 20% more CO2 than we thought, new models find

Posted on by

[…]

Using realistic ecological modeling, scientists led by Western Sydney University’s Jürgen Knauer found that the globe’s vegetation could actually be taking on about 20% more of the CO2 humans have pumped into the atmosphere and will continue to do so through to the end of the century.

“What we found is that a well-established climate model that is used to feed into global climate assessments by the likes of the IPCC (Intergovernmental Panel on Climate Change) predicts stronger and sustained carbon uptake until the end of the 21st century when extended to account for the impact of some critical physiological processes that govern how plants conduct photosynthesis,” said Knauer.

[…]

Current models, the team adds, are not that complex so likely underestimate future CO2 uptake by vegetation.

[…]

Taking the well-established Community Atmosphere-Biosphere Land Exchange model (CABLE), the team accounted for three physiological factors […] the team found that the most complex version, which accounted for all three factors, predicted the most CO2 uptake, around 20% more than the simplest formula.

[…]

“Our understanding of key response processes of the carbon cycle, such as plant photosynthesis, have advanced dramatically in recent years,” said Ben Smith, professor and research director of Western Sydney University’s Hawkesbury Institute for the Environment. “It always takes a while for new knowledge to make it into the sophisticated models we rely on to inform climate and emissions policy. Our study demonstrates that by fully accounting for the latest science in these models can lead to materially different predictions.

[…]

And while it’s somewhat good news, the team says plants can’t be expected to do all the heavy lifting; the onus remains on governments to stick to emission reduction obligations. However, the modeling makes a strong case for the value of greening projects and their importance in comprehensive approaches to tackling global warming.

[…]

Source: Plants may be absorbing 20% more CO2 than we thought, new models find

HP’s CFO tells world: we are locking in customers for more profit

Posted on by

[…] Tech vendors – software, hardware, and cloud services – generally avoid terms that suggest they’re perhaps in some way pinning down customers in a strategic sales hold.

But as Marie Myers, chief financial officer at HP, was this week talking to the UBS Global Technology conference, in front of investors, the thrust of the message was geared toward the audience.

“We absolutely see when you move a customer from that pure transactional model … whether it’s Instant Ink, plus adding on that paper, we sort of see a 20 percent uplift on the value of that customer because you’re locking that person, committing to a longer-term relationship.”

Instant Ink is a subscription in which ink or toner cartridges are dispatched when needed, with customers paying for plans that start at $0.99 and run to $25.99 per month. As of May last year, HP had more than 11 million subscribers to the service. Since then it has banked double-digit percentage figures on the revenues front.

By pre-pandemic 2019, HP had grown weary of third-party cartridge makers stealing its supplies business. It pledged to charge more upfront for certain printer hardware (“rebalance the system profitability, capturing more profit upfront”).

HP also set in motion new subscriptions, and launched Smart Tank hardware filled with a pre-defined amount of ink/toner. These now account for 60 percent of total shipments.

Myers told the UBS Conference she was “really proud” that HP could “raise the range on our print margins” based on “bold moves and shifting models.”

[…]

An old industry factoid from 2003 was that HP ink cost seven times more than a bottle of 1985 Dom Perignon. HP isn’t alone in these sorts of comparisons – Epson was called out by Which? a couple years back.

[…]

Source: Vendor lock-in is a good thing? HP’s CFO thinks so