Technology

cross-posted from: https://jlai.lu/post/10771034

Personal review:

A good recap of his previous writings and talks on the subject for the first third, but a bit long. Having paid attention to them for the past year or two, my attention started drifting a few times. I ended up being more impressed with how much he's managed to condense explaining "enshittification" from 45+ minutes down to around 15.

As soon as he starts building off of that to work towards the core of his message for this talk, I was more-or-less glued to the screen. At first because it's not exactly clear where he's going, and there are (what felt like) many specific court rulings to keep up with. Thankfully, once he has laid enough groundwork he gets straight his point. I don't want to spoil or otherwise lessen the performance he gives, so I won't directly comment on what his point is in the body of this post - I think the comments are better suited for that anyways.

I found the rest to be pretty compelling. He rides the fine line between directionless discontent and overenthusiastic activist-with-a-plan as he doubles down on his narrative by calling back to the various bits of groundwork he laid before - now that we're "in" on the idea, what felt like stumbling around in the dark turns into an illuminating path through some of the specifics of the last twenty to forty years of the dynamics of power between tech bosses and their employees. The rousing call to action was also great way to end and wrap it all up.

I've become very biased towards Cory Doctorow's ideas, in part because they line up with a lot of the impressions I have from my few years working as a dev in a big-ish multinational tech company. This talk has done nothing to diminish that bias - on the contrary.

WASHINGTON, Sept 21 (Reuters) - The U.S. Commerce Department is expected on Monday to propose prohibiting Chinese software and hardware in connected and autonomous vehicles on American roads due to national security concerns, two sources told Reuters.

The Biden administration has raised serious concerns about the collection of data by Chinese companies on U.S. drivers and infrastructure as well as the potential foreign manipulation of vehicles connected to the internet and navigation systems.

The proposed regulation would ban the import and sale of vehicles from China with key communications or automated driving system software or hardware, said the two sources, who declined to be identified because the decision had not been publicly disclosed.

The move is a significant escalation in the United States' ongoing restrictions on Chinese vehicles, software and components. Last week, the Biden administration locked in steep tariff hikes on Chinese imports, including a 100% duty on electric vehicles as well as new hikes on EV batteries and key minerals.

LONDON, Sept 20 (Reuters) - The world's biggest technology companies have embarked on a final push to persuade the European Union to take a light-touch approach to regulating artificial intelligence as they seek to fend off the risk of billions of dollars in fines.

EU lawmakers in May agreed the AI Act, the world's first comprehensive set of rules governing the technology, following months of intense negotiations between different political groups.

https://ghostarchive.org/varchive/iRv_syz2DAc

Alibaba Cloud, the cloud computing arm of China’s Alibaba Group Ltd., today announced the release of more than 100 new artificial intelligence large language models open source as part of the Qwen 2.5 family of models.

Revealed at the company’s Apsara Conference, the new model series follows the release of the company’s foundation model Tongyi Qianwen, or Qwen, last year. Since then, the Qwen models have been downloaded more than 40 million times across platforms such as Hugging Face and Modelscope.

The new models range from sizes as small as a half-billion parameters to as large as 72 billion parameters. In an LLM, parameters define the behavior of an AI model and what it uses to make predictions about its skills such as mathematics, coding or expert knowledge.

Smaller, more lightweight models can be trained quickly using far less processing power on more focused training sets and excel at simpler tasks. In contrast, larger models need heavy processing power and longer training times and generally perform better on complex tasks requiring deep language understanding.

cross-posted from: https://lemm.ee/post/42612055

(Asking for the civilized world.)

Google won a legal challenge Wednesday against a €1.49 billion ($1.66 billion) antitrust fine from the European Union, while chipmaker Qualcomm failed to repeal a penalty.

The rulings underscore the mixed record of outgoing EU antitrust chief Margrethe Vestager in defending her crackdown on Big Tech in court. She scored two major wins last week: against Google in a separate case and against Apple’s tax deal with Irish authorities.

In a 2019 decision, the European Commission said Google, owned by Alphabet (GOOGL), had abused its dominance to prevent websites from using brokers other than its AdSense platform that provided search ads. The practices that it said were illegal took place from 2006 to 2016.

Apple has rolled out macOS Sequoia as its latest software updates for supported Mac computers, along with iPadOS 18, watchOS 11 and tvOS 18. These updates introduce new features and customisation options on Apple's devices, while future updates will add support for Apple Intelligence — new features that are powered by artificial intelligence (AI) on the iPhone 15 Pro models, the new iPhone 16 lineup, and select tablets and Mac models with Apple's M-series chipsets. These software updates were first shown off at Apple's Worldwide Developers Conference (WWDC 2024).

Here are some of the most notable changes on macOS Sequoia, iOS 18, watchOS 11, tvOS 18, and visionOS 2. You can update eligible devices to these software versions, but it's better to back up important data before installing major operating system updates. You can also read more about Apple's iOS 18 update that is rolling out to eligible smartphones.

https://archive.is/2024.09.16-214410/https://interestingengineering.com/innovation/china-starlink-detect-stealth-fighters#selection-373.0-373.86

SAN FRANCISCO, Sept 16 (Reuters) - Intel (INTC.O) lost out on a contract to design and fabricate Sony’s PlayStation 6 chip in 2022, which dealt a significant blow to its effort to build its fledgling contract manufacturing business, according to three sources with knowledge of the events.

The effort by Intel to win out over Advanced Micro Devices (AMD.O)in a competitive bidding process to supply the design for the forthcoming PlayStation 6 chip and Taiwan Semiconductor Manufacturing Co (2330.TW) as the contract manufacturer would have amounted to billions of dollars of revenue and fabricating thousands of silicon wafers a month, two sources said.

Intel and AMD were the final two contenders in the bidding process for the contract.

Winning the Sony (6758.T) PlayStation 6 chip design business would have been a victory for Intel's design segment and would have doubled as a win for the company's contract manufacturing effort, or foundry business, which was the centerpiece of Intel CEO Pat Gelsinger’s turnaround plan.

Mark Zuckberg’s Meta is to go ahead with controversial plans to use millions of UK Facebook and Instagram posts to train its artificial intelligence (AI) technology, in a practice that is effectively outlawed under EU privacy laws.

Meta said it had “engaged positively” with the Information Commissioner’s Office (ICO) over the plan, after it paused similar proposals in June in the UK and EU. The pause came after the ICO warned tech firms to respect the privacy of users when building generative AI.

On Friday, the ICO made it clear it has not provided regulatory approval for the plan, but will instead monitor the experiment after Meta agreed changes to its approach. These include making it easier for users to opt out of allowing their posts to be processed for AI.