Month: November 2024

Die Sicherheit sensibler Daten ist für Unternehmen von essenzieller Bedeutung. Der dreitägige Online-Workshop der Golem Karrierewelt bietet eine Einführung in die IT-Grundschutz-Methodik des BSI - inklusive Zertifikatsprüfung. (Golem Karrierewelt, Server-Applikationen)

An Asian elephant named Mary living at the Berlin Zoo surprised researchers by figuring out how to use a hose to take her morning showers, according to a new paper published in the journal Current Biology. “Elephants are amazing with hoses,” said co-author Michael Brecht of the Humboldt University of Berlin. “As it is often the case with elephants, hose tool use behaviors come out very differently from animal to animal; elephant Mary is the queen of showering.”

Tool use was once thought to be one of the defining features of humans, but examples of it were eventually observed in primates and other mammals. Dolphins have been observed using sea sponges to protect their beaks while foraging for food, and sea otters will break open shellfish like abalone with rocks. Several species of fish also use tools to hunt and crack open shellfish, as well as to clear a spot for nesting. And the coconut octopus collects coconut shells, stacking them and transporting them before reassembling them as shelter.

Birds have also been observed using tools in the wild, although this behavior was limited to corvids (crows, ravens, and jays), although woodpecker finches have been known to insert twigs into trees to impale passing larvae for food. Parrots, by contrast, have mostly been noted for their linguistic skills, and there has only been limited evidence that they use anything resembling a tool in the wild. Primarily, they seem to use external objects to position nuts while feeding.

Read full article

Comments

On Friday, research organization Epoch AI released FrontierMath, a new mathematics benchmark that has been turning heads in the AI world because it contains hundreds of expert-level problems that leading AI models solve less than 2 percent of the time, according to Epoch AI. The benchmark tests AI language models (such as GPT-4o, which powers ChatGPT) against original mathematics problems that typically require hours or days for specialist mathematicians to complete.

FrontierMath's performance results, revealed in a preprint research paper, paint a stark picture of current AI model limitations. Even with access to Python environments for testing and verification, top models like Claude 3.5 Sonnet, GPT-4o, o1-preview, and Gemini 1.5 Pro scored extremely poorly. This contrasts with their high performance on simpler math benchmarks—many models now score above 90 percent on tests like GSM8K and MATH.

The design of FrontierMath differs from many existing AI benchmarks because the problem set remains private and unpublished to prevent data contamination. Many existing AI models are trained on other test problem datasets, allowing the AI models to easily solve the problems and appear more generally capable than they actually are. Many experts cite this as evidence that current large language models (LLMs) are poor generalist learners.

Read full article

Comments

Barely nine months after the last cut, NASA's Jet Propulsion Laboratory will again reduce its workforce. On Wednesday, the lab will lay 325 employees off, representing about 5 percent of the workforce at the California-based laboratory that leads the development of robotic space probes for NASA.

"This is a message I had hoped not to have to write," JPL Director Laurie Leshin said in a memo to staff members on Tuesday morning, local time. "Despite this being incredibly difficult for our community, this number is lower than projected a few months ago thanks in part to the hard work of so many people across JPL."

The cuts this week follow a reduction of 530 employees in February of this year due to various factors, including a pause in funding for the Mars Sample Return mission. The NASA laboratory has now cut about one-eighth of its workforce this year.

Read full article

Comments

For years now, many AI industry watchers have looked at the quickly growing capabilities of new AI models and mused about exponential performance increases continuing well into the future. Recently, though, some of that AI "scaling law" optimism has been replaced by fears that we may already be hitting a plateau in the capabilities of large language models trained with standard methods.

A weekend report from The Information effectively summarized how these fears are manifesting amid a number of insiders at OpenAI. Unnamed OpenAI researchers told The Information that Orion, the company's codename for its next full-fledged model release, is showing a smaller performance jump than the one seen between GPT-3 and GPT-4 in recent years. On certain tasks, in fact, the upcoming model "isn't reliably better than its predecessor," according to unnamed OpenAI researchers cited in the piece.

On Monday, OpenAI co-founder Ilya Sutskever, who left the company earlier this year, added to the concerns that LLMs were hitting a plateau in what can be gained from traditional pre-training. Sutskever told Reuters that "the 2010s were the age of scaling," where throwing additional computing resources and training data at the same basic training methods could lead to impressive improvements in subsequent models.

Read full article

Comments

The big three record labels notched another court victory against a broadband provider last month, but the music publishing firms aren't happy that an appeals court only awarded per-album damages instead of damages for each song.

Universal, Warner, and Sony are seeking an en banc rehearing of the copyright infringement case, claiming that Internet service provider Grande Communications should have to pay per-song damages over its failure to terminate the accounts of Internet users accused of piracy. The decision to make Grande pay for each album instead of each song "threatens copyright owners' ability to obtain fair damages," said the record labels' petition filed last week.

The case is in the conservative-leaning US Court of Appeals for the 5th Circuit. A three-judge panel unanimously ruled last month that Grande, a subsidiary of Astound Broadband, violated the law by failing to terminate subscribers accused of being repeat infringers. Subscribers were flagged for infringement based on their IP addresses being connected to torrent downloads monitored by Rightscorp, a copyright-enforcement company used by the music labels.

Read full article

Comments

Bitcoin hit a new record high late Monday, its value peaking at $89,623 as investors quickly moved to cash in on expectations that Donald Trump will end a White House crackdown that intensified last year on crypto.

While the trading rally has now paused, analysts predict that bitcoin's value will only continue rising following Trump's win—perhaps even reaching $100,000 by the end of 2024, CNBC reported.

Bitcoin wasn't the only winner emerging from the post-election crypto trading. Crypto exchanges like Coinbase also experienced surges in the market, and one of the biggest winners, CNBC reported, was dogecoin, a cryptocurrency linked to Elon Musk, who campaigned for Trump and may join his administration. Dogecoin's value is up 135 percent since Trump's win.

Read full article

Comments

For most of 2024, major football leagues in Spain and Italy have regularly used the media to criticize and then publicly threaten Google. Rightsholders have always criticized Google but this feels a bit different.

More recently, flare-ups have coincided with rightsholders demanding that Google implements new, controversial anti-piracy measures, such as remotely deleting apps from users’ phones. Whether any discussions take place in private is unclear. But from the outside, subsequent one-sided criticism of Google’s response to piracy via the media, suggests that differences still exist.

Recent Allegations

Criticism rose again via Italian media during October, with headlines such as ‘TV rights and piracy, Serie A ready to sue Google‘ and variations thereof. The common theme in these reports are allegations that Google receives takedown notices from football league Serie A but then ignores them; all part of a reluctance to collaborate on anti-piracy matters in general, the allegations claim.

To a background of Google receiving its 10 billionth URL removal request just over a week ago, and limited complaints along similar lines from elsewhere, ignoring takedown notices seems unlikely. At the very least, it’s lacking context in the bigger picture.

Transparency Helps to Settle Disputes

Details of specific complaints haven’t been made public but multiple reports state that Serie A’s complaints relate to Google’s alleged failures in the current season, which started mid-August. The table below shows takedown notices filed by Serie A or appointed agents between May 26 and September 23, 2024.

Serie A appears to use more than one account when filing takedown notices, including Lega Nazionale Professionisti Serie A and Liga Nacional de Fütbol Profesional, which we’ll return to in a moment.

The complaint at the bottom of the table dated May 26 was sent before the season began but is relevant in the bigger picture. It requested the removal of 84 URLs, of which Google deleted 12% from search results. Since the remaining 88% of URLs in the notice did not actually exist in Google’s search indexes when the notice was sent, removal was impossible.

Google Processes Unnecessarily Confusing Complaints

Here’s where things become unnecessarily complicated. The May 26 takedown notice was sent by an agent; more specifically a company called Sportian, the anti-piracy company previously known as LaLiga Tech, owned by top tier Spanish football league, LaLiga. The full takedown notice, courtesy of the Lumen Database, can be viewed here with a sample of the not-in-index URLs shown below.

At the time of writing, a total of 3,530 individual requests filed with Google under the Liga Nacional de Fútbol Profesional account have requested the removal of 437,524 URLs. Google went on to remove 49.8% of that total, could not remove 23.3% because they did not exist in its indexes and a further 3.3% because the same URLs were duplicates of previously reported links.

The issue here is that Liga Nacional de Fútbol Profesional isn’t a reference to Serie A; it’s actually Sportiva using the full name of Spanish football league LaLiga in its unlikely capacity as a takedown agent for Serie A.

While there doesn’t appear to be anything nefarious going on here, takedown fraud thrives on convoluted situations not dissimilar to this. If given time to do its job, Google often identifies abuses and prevents them from doing damage.

Despite a hall of mirrors, Google still processed the notices and also took down whatever it was able to take down. According to Google data on all of Serie A accounts mentioned earlier, 70.2% of all URLs requested for takedown do not exist in its search indexes so are impossible to delete.

Season Begins With a Change of Tactics

The Serie A notices dated Aug 22, Aug 27, and September 3, were not sent to Google from its usual accounts. Instead, they were filed by SP Tech, the law firm/anti-piracy company behind the Piracy Shield platform.

Those three notices contain just 12 URLs in total, of which Google removed none.

All three takedown notices

After sending takedown notices listing specific URLs and infringing content as required under the DMCA, there’s a very clear switch here to domain-based takedown notices that appear to identify no infringing content at all.

Indeed, while these domains appear to link to sites that may well offer IPTV packages containing Serie A matches, the DMCA states that takedown notices must contain sufficient information for the recipient to identify and remove the infringing content.

Clash of Legal Requirements

Under Law n. 93 of 14 July 2023, which has been in force in Italy since August 8, 2023, service providers and ISPs are required to block or otherwise prevent access to infringing content within 30 minutes of receiving an instruction from Piracy Shield; a system to which Google is not currently connected.

Even if Google deindexed the domains in Serie A’s notices, that wouldn’t prevent access to any IPTV streams. That’s something this law does not consider. These are blocking and deindexing orders made with no judicial oversight, that have shown to be erroneous on several occasions, with Google itself blocked in error a matter of weeks ago.

The exact reasons for Google currently rejecting domain-only takedown/deindexing demands are unknown. In the UK and the Netherlands, Google has been deindexing entire pirate sites since 2021. The difference is that rightsholders in both countries previously obtained court orders that provide a legal basis for ISP blocking. Google isn’t even mentioned in those orders, but it cooperates because due process requirements have been met.

In Italy, purely on the word of a rightsholder, who may or may not know the difference between Google Drive and a pirate site on any given day, companies like Google are expected to blindly and immediately follow orders. Thirty minutes is a timeframe that encourages no checks whatsoever, the consequences of which have been widely publicized.

Called Out in Public

Resistance, it appears, can lead to blanket accusations of doing almost nothing to assist rightsholders. As highlighted above, Google does process takedown notices; however, depending on the takedown account held up as proof, Google can be shown to be mostly or even totally non-compliant. The Sp Tech account shown below, which was introduced only recently, is one such example.

Describing Google as generally non-compliant, or even “grossly negligent” according to a recent comment, simply isn’t true. Serie A has another account at Google that’s been in use since January 2019. Over more than five years, Serie A filed 56,847 individual takedown notices that together requested the removal of 1,194,826 URLs.

Of that total, 65.5% of the URLs did not exist in Google search, meaning it was impossible to remove them. Google removed 27.9% of the URLs as requested, leaving duplicate URLs already dealt with and 5.4% for which Google took no action.

While that sounds like a minimum compliance rate of 95%, attention will inevitably focus on the remaining 5% and why those weren’t removed as well. There are many reasons behind Google’s refusals to take action, including protecting innocent parties from abusive or simply careless takedown demands.

That Google continues to do that, even while being publicly disparaged for its apparent failings, is commendable. That it prevented Serie A from deindexing its own website on October 27, is ironic, to put it mildly.

From: TF, for the latest news on copyright battles, piracy and more.

Spotify has lost all enthusiasm for the little music devices it sold for just half a year. Firmware hackers, as usually happens, have a lot more interest and have stepped in to save, and upgrade, a potentially useful gadget.

Spotify's idea a couple years ago was a car-focused device for those who lacked Apple CarPlay, Android Auto, or built-in Spotify support in their vehicles, or just wanted a dedicated Spotify screen. The Car Thing was a $100 doodad with a 4-inch touchscreen and knob that attached to the dashboard (or into a CD slot drive). All it could do was play Spotify, and only if you were a paying member, but that could be an upgrade for owners of older cars, or people who wanted a little desktop music controller.

But less than half a year after it fully released its first hardware device, Spotify gave up on the Car Thing due to "several factors, including product demand and supply chain issues." A Spotify rep told Ars that the Car Thing was meant "to learn more about how people listen in the car," and now it was "time to say goodbye to the devices entirely." Spotify indicated it would offer refunds, though not guaranteed, and moved forward with plans to brick the device in December 2024.

Read full article

Comments

In some ways, my review of the new MacBook Pros will be a lot like my review of the new iMac. This is the third year and fourth generation of the Apple Silicon-era MacBook Pro design, and outwardly, few things have changed about the new M4, M4 Pro, and M4 Max laptops.

Here are the things that are different. Boosted RAM capacities, across the entire lineup but most crucially in the entry-level $1,599 M4 MacBook Pro, make the new laptops a shade cheaper and more versatile than they used to be. The new nano-texture display option, a $150 upgrade on all models, is a lovely matte-textured coating that completely eliminates reflections. There's a third Thunderbolt port on the baseline M4 model (the M3 model had two), and it can drive up to three displays simultaneously (two external, plus the built-in screen). There's a new webcam. It looks a little nicer and has a wide-angle lens that can show what's on your desk instead of your face if you want it to. And there are new chips, which we'll get to.

That is essentially the end of the list. If you are still using an Intel-era MacBook Pro, I'll point you to our previous reviews, which mostly celebrate the improvements (more and different kids of ports, larger screens) while picking one or two nits (they are a bit larger and heavier than late-Intel MacBook Pros, and the display notch is an eyesore).

Read full article

Comments