Saturday, August 26, 2023
Friday, August 25, 2023
Monday, August 21, 2023
Elon Musk admits X ‘may fail’ after glitch deletes Twitter photos
Elon Musk admits X ‘may fail’ after glitch deletes Twitter photos
“Platform’s users voice concern as it removes pictures and links on posts made before December 2014
Elon Musk, the owner of the app formerly known as Twitter, has said the social media site “may fail”, after a glitch caused pictures posted before December 2014 to be deleted.
In a post on the site, renamed X, Musk said: “The sad truth is that there are no great ‘social networks’ right now.”
He added: “We may fail, as so many have predicted, but we will try our best to make there be at least one.”
Over the weekend, a glitch on the platform meant that the site removed pictures and links on posts made before December 2014. The posts showed broken links instead of the pictures and videos that were previously there.
Several users noticed the glitch, with the technologist Tom Coates among those pointing it out. Coates referred to the glitch as “epic vandalism by Musk” and suggested it could be a cost-saving exercise.
Ellen DeGeneres’ famous Oscars selfie from 2014 was also removed but later restored. The photo became the platform’s most retweeted photo ever, with more than 2m shares on the platform. Former US president Barack Obama’s viral tweet after his 2012 election win remained unaffected.
Some users speculated that the glitch was caused by an effort to save costs on storage data, while others attributed it to changes made on the platform in 2016 when it added “enhanced URL enrichment”, designed to show previews for linked websites and attachments beyond the company’s 140 character limit, according to The Verge.
The malfunction came after reports last week that suggested access had been slowed down from X to other social media sites including the Meta platforms Facebook, Threads and Instagram. The five-second delay also appeared on links to some news sites such as the New York Times and Reuters.
Earlier this year, users noticed that they could no longer post or send messages to each other, and saw a notification that said they were “over the daily limit for sending tweets”. Direct messages were also not working. The company apologised for “the trouble” in a post. Previous glitches have also left users unable to log in or view their own posts.
Twitter has cut thousands of jobs since Musk’s takeover, vastly reducing the workforce since November.
Despite the Tesla chief executive’s aggressive efforts to cut down costs of the platform since his takeover, he has faced challenges with revenue. Last month, he reported a 50% drop in advertising revenue, along with a heavy debt load. The platform faces annual interest payments that stand at $1.5bn as a result of the debt it took on as part of last October’s $44bn deal that turned the company private.
According to the Wall Street Journal, the social media company also faces suits claiming more than $14m (£11m) in unpaid bills since Musk took over.
X was contacted for comment.
I hope you appreciated this article. Before you move on, I was hoping you would consider taking the step of supporting the Guardian’s journalism.
From Elon Musk to Rupert Murdoch, a small number of billionaire owners have a powerful hold on so much of the information that reaches the public about what’s happening in the world. The Guardian is different. We have no billionaire owner or shareholders to consider. Our journalism is produced to serve the public interest – not profit motives.
And we avoid the trap that befalls much US media – the tendency, born of a desire to please all sides, to engage in false equivalence in the name of neutrality. While fairness guides everything we do, we know there is a right and a wrong position in the fight against racism and for reproductive justice. When we report on issues like the climate crisis, we’re not afraid to name who is responsible. And as a global news organization, we’re able to provide a fresh, outsider perspective on US politics – one so often missing from the insular American media bubble.
Around the world, readers can access the Guardian’s paywall-free journalism because of our unique reader-supported model. That’s because of people like you. Our readers keep us independent, beholden to no outside influence and accessible to everyone – whether they can afford to pay for news, or not.“
AI-generated art cannot be copyrighted, rules a US Federal Judge
AI-generated art cannot be copyrighted, rules a US Federal Judge
:format(webp)/cdn.vox-cdn.com/uploads/chorus_asset/file/13292779/acastro_181017_1777_brain_ai_0002.jpg)
Illustration by Alex Castro / The Vergenone
“United States District Court Judge Beryl A. Howell ruled on Friday that AI-generated artwork can’t be copyrighted, as noted by The Hollywood Reporter. She was presiding over a lawsuit against the US Copyright Office after it refused a copyright to Stephen Thaler for an AI-generated image made with the Creativity Machine algorithm he’d created.
Thaler had tried multiple times to copyright the image “as a work-for-hire to the owner of the Creativity Machine,” which would have listed the author as the creator of the work and Thaler as the artwork’s owner, but he was repeatedly rejected.
After the Office’s final rejection last year, Thaler sued the Office, claiming its denial was “arbitrary, capricious ... and not in accordance with the law,” but Judge Howell didn’t see it that way. In her decision, Judge Howell wrote that copyright has never been granted to work that was “absent any guiding human hand,” adding that “human authorship is a bedrock requirement of copyright.”
That’s been borne out in past cases cited by the judge, like that one involving a monkey selfie. To contrast, Judge Howell noted a case in which a woman compiled a book from notebooks she’d filled with “words she believed were dictated to her” by a supernatural “voice” was worthy of copyright.
Judge Howell did, however, acknowledge that humanity is “approaching new frontiers in copyright,” where artists will use AI as a tool to create new work. She wrote that this would create “challenging questions regarding how much human input is necessary” to copyright AI-created art, noting that AI models are often trained on pre-existing work.
Stephen Thaler plans to appeal the case. His attorney, Ryan Abbot of Brown Neri Smith & Khan LLP, said, “We respectfully disagree with the court’s interpretation of the Copyright Act,” according to Bloomberg Law, which also reported a US Copyright Office statement saying it believed the court’s decision was the right one.
Nobody really knows how things will shake out around US copyright law and artificial intelligence, but the court cases have been piling up. Sarah Silverman and two other authors filed suit against OpenAI and Meta earlier this year over their models’ data scraping practices, for instance, while another lawsuit by programmer and lawyer Matthew Butterick alleges that data scraping by Microsoft, GitHub, and OpenAI amounted to software piracy.“
Twitter held in contempt, fined $350K over Trump data delay | Ars Technica
Twitter held in contempt, fined $350K over Trump data delay
Trump accused the DOJ of "secretly attacking" his Twitter account.
Today, an unsealed court document revealed that, earlier this year, a federal judge held Twitter (now called X) in contempt of court. The judge imposed $350,000 in sanctions.
Sanctions were applied after the social media platform delayed compliance with a federal search warrant that required Twitter to hand over Donald Trump's Twitter data without telling the former president about the warrant for 180 days.
Enter your email to get the Ars Technica newsletter
At first, Twitter resisted producing Trump's data and argued that the government's nondisclosure order violated the First Amendment and the Stored Communications Act. However, US circuit judge Florence Pan wrote that the court was largely unpersuaded by Twitter's arguments, mostly because the government's interest in Trump's data as part of its ongoing January 6 investigation was "unquestionably compelling."
Last week, Trump was indicted on felony charges for working to overturn the results of the 2020 election, AP reported. Trump has since pled not guilty and continues to post on Truth Social, where he commented on today's Twitter revelations. The former president said that he "just found out" that "crooked" Joe Biden's Justice Department "secretly attacked" his Twitter account. He called the search warrant a "major 'hit'" to his civil rights.
"Does the First Amendment still exist?" Trump wrote.
Twitter did not respond to Ars' request to comment.
Why was Twitter held in contempt?
The unsealed court document provides a timeline of events leading the court to hold Twitter in contempt.
Twitter's troubles started on January 17, 2023, when the government secured a search warrant that "directed Twitter to produce data and records related to the '@realDonaldTrump' Twitter account." The government then took the extra step to apply for a nondisclosure order, which was granted because "the district court found that there were 'reasonable grounds to believe' that disclosing the warrant to former President Trump 'would seriously jeopardize the ongoing investigation' by giving him 'an opportunity to destroy evidence, change patterns of behavior, [or] notify confederates.'"
The government immediately tried to serve Twitter with the search warrant—which required Trump's data to be shared within 10 days—but the website where Twitter gathers legal requests was "inoperative." It took two days before Twitter's website was fixed and the government was able to submit the warrant, and even then, Twitter did not immediately respond. Finally, on January 25, the government reached Twitter's counsel, who said "that she 'had not heard anything about [the] [w]arrant.'"
Then, on February 1, four days after failing to meet the deadline, Twitter "objected to producing any of the account information." The government immediately requested a hearing where Twitter had to show cause why it should not be held in contempt for missing the deadline.
That hearing was scheduled for February 7. During the hearing, Twitter "requested that the court stay its enforcement of the warrant until after it adjudicated Twitter's motion to vacate or modify the nondisclosure order." Essentially, Twitter didn't want to hand over data without Trump's knowledge, but the court denied Twitter's request and "found Twitter in contempt of court."
At that point, Twitter was given an "opportunity to purge its contempt by producing" Trump's account information. The court checked with Twitter and confirmed that it was capable of meeting a rapid deadline and turning over the data by 5:00 pm that evening.
"I believe we are prepared to do that," Twitter's counsel said. "Yes, Your Honor."
Twitter missed the deadline, again
To encourage Twitter to meet this new deadline, the court threatened hefty fines for non-compliance. The judge laid out a formula for "sanctions that would accrue at a geometric rate: $50,000 per day, to double every day that Twitter did not comply." At that time, Twitter did not object to the sanctions formula, Pan noted in the newly unsealed opinion.
"Twitter missed the 5pm deadline," the document said. "Although Twitter timely produced some records, its production was incomplete."
Ultimately, it took Twitter until 8:06 pm on February 9 to produce all the data that the government sought. According to the court, that was three days late. Twitter tried and failed to argue its way out of sanctions by saying that it acted in good faith to supply information as quickly as possible.
"Twitter contends that it 'substantially complied with the [w]arrant' because 'there was nothing [it] could have done to comply faster' after the court issued the February 7 order," the court document said.
The court rejected Twitter's "good faith" arguments, mainly because the company blew past the original deadline and repeatedly failed to raise concerns at earlier opportunities.
On March 3, the court denied Twitter's attempt to modify or vacate the nondisclosure order, found Twitter in civil contempt, and imposed the $350,000 contempt sanction.
While Twitter appealed the decision, the company "paid the $350,000 sanction into an escrow account maintained by the district court clerk's office."
The appeal process stretched on for months. On June 20, the government modified its nondisclosure order to "permit Twitter to notify the former President of the existence and contents of the warrant." That meant that "the only limitation on the disclosure would be" for Twitter "to withhold the identity of the case agent assigned to the investigation."
Meanwhile, Twitter was late in its attempts to oppose the sanctions formula. The court opinion said that Twitter's counsel "belatedly" pointed out that "roughly one month of noncompliance" would have "required Twitter to pay a sanction greater than 'the entire world's gross domestic product.'"
Twitter continued challenging the nondisclosure order and the sanctions, but the court rejected most of its arguments and ultimately affirmed the contempt sanctions, issuing its opinion on July 18.
"Twitter never raised any objection to the sanctions formula, despite having several opportunities to do so," Pan wrote in the opinion. "The company thus appeared to acquiesce to the formula. Moreover, the $350,000 sanction ultimately imposed was not unreasonable, given Twitter's $40 billion valuation and the court's goal of coercing Twitter's compliance."
Sunday, August 20, 2023
Physicist Michio Kaku: ‘We could unravel the secrets of the universe’
Physicist Michio Kaku: ‘We could unravel the secrets of the universe’
“Have you been feeling anxious about technology lately? If so, you’re in good company. The United Nations has urged all governments to implement a set of rules designed to rein in artificial intelligence. An open letter, signed by such luminaries as Yuval Noah Harari and Elon Musk, called for research into the most advanced AI to be paused and measures taken to ensure it remains “safe … trustworthy, and loyal”. These pangs followed the launch last year of ChatGPT, a chatbot that can write you an essay on Milton as easily as it can generate a recipe for everything you happen to have in your cupboard that evening.
But what if the computers used to develop AI were replaced by ones able to make calculations not millions, but trillions of times faster? What if tasks that might take thousands of years to perform on today’s devices could be completed in a matter of seconds? Well, that’s precisely the future that physicist Michio Kaku is predicting. He believes we are about to leave the digital age behind for a quantum era that will bring unimaginable scientific and societal change. Computers will no longer use transistors, but subatomic particles, to make calculations, unleashing incredible processing power. Another physicist has likened it to putting “a rocket engine in your car”. How are you feeling now?
Kaku seems pretty relaxed about it all – some might say boosterish. He talks to me via Zoom from his apartment on Manhattan’s Upper West Side. Seventy-six and retired from research, he still teaches at the City University of New York where he is professor of theoretical physics and gets to do the fun stuff. A fan of Isaac Asimov, he tells me that he’s currently teaching a course on the physics of science fiction. “I talk about what is known and not known about time travel, space warps, the multiverse, all the things you see in Marvel Comics, I break it down.” His website describes him as a “futurist and populariser of science” and his new book, Quantum Supremacy, sketches out all the promise of quantum computing and very little of the downside. Though he has the long white hair of the stereotypical mad scientist, it is swept back elegantly. He speaks at the pace of a practised lecturer, with the occasional outbreak of mild bemusement pitching his voice a little higher.
Kaku has a simple explanation for the doom-mongering around ChatGPT: “Journalists are hyperventilating about chatbots … because they see that their job is on the line. Many jobs have been on the line historically, but no one really said much about them. Now, journalists are right there in the crosshairs.” This is a somewhat partial view – a report by Goldman Sachs recently estimated that 300m jobs are at risk of automation as a result of AI. Kaku does admit that we might see “sentient machines” emerging from laboratories but reckons that could take another hundred years or so. In the meantime, he thinks there’s a lot to feel good about.
The rocket engine of quantum computing will, Kaku says, completely transform research in chemistry, biology and physics, with all sorts of knock-on effects. Among other things, it will enable us to take CO2 out of the atmosphere and turn it into fuel, with the waste products captured and used again – so-called carbon recycling. It will help us extract nitrogen from the air without the high temperatures and pressures that mean fertiliser production currently accounts for 2% of the energy used on Earth, leading to a new green revolution. It will allow us to create super-efficient batteries to help renewables go further (today’s lithium-ion batteries only carry about 1% of the energy stored in gasoline). It will solve the design and engineering challenges currently stopping us from generating cheap, abundant power via nuclear fusion. And it will lead to radically effective treatments for cancer, Alzheimer’s and Parkinson’s diseases, alongside a host of others.
How? The main thing to understand is that quantum computers can make calculations much, much faster than digital ones. They do this using qubits, the quantum equivalent of bits – the zeros and ones that convey information in a conventional computer. Whereas bits are stored as electrical charges in transistors etched on to silicon chips, qubits are represented by properties of particles, for example, the angular momentum of an electron. Qubits’ superior firepower comes about because the laws of classical physics do not apply in the strange subatomic world, allowing them to take any value between zero and one, and enabling a mysterious process called quantum entanglement, which Einstein famously called spukhafte Fernwirkung or “spooky action at a distance”. Kaku makes valiant efforts to explain these mechanisms in his book, but it’s essentially impossible for a layperson to fully grasp. As the science communicator Sabine Hossenfelder puts it in one of her wildly popular YouTube videos on the subject: “When we write about quantum mechanics, we’re faced with the task of converting mathematical expressions into language. And regardless of which language we use, English, German, Chinese or whatever, our language didn’t evolve to describe quantum behaviour.”
What we’re left with are analogies of varying helpfulness, for example the toy trains with compasses on them and mice in mazes that Kaku invokes to explain such complex ideas as superposition and path integrals. Beyond these, there is one important takeaway: reality is quantum, and so quantum computers can simulate it in a way that digital ones struggle to. “Mother Nature does not compute digitally,” he tells me. “Quantum computers should [be able to] unravel the secrets of life, the secrets of the universe, the secrets of matter, because the language of nature is the quantum principle.” If you want to know precisely how photosynthesis works (still a mystery to modern science), or how one protein interacts with another in the human body, you will be able to use the “virtual lab” of a quantum computer to model it precisely. Designing medicines to interrupt biological processes gone awry, like the proliferation of cancer cells or the misfolding of proteins in Alzheimer’s disease, could become much easier. Kaku even reckons that the riddle of ageing will be unravelled so that we can arrest it – one of the chapters in his book is called simply “Immortality”.
At this stage, it’s worth introducing an important caveat. Quantum computers are very, very hard to make. Because they rely on tiny particles that are extremely sensitive to any kind of disturbance, most can only run at temperatures close to absolute zero, where everything slows down and there’s minimal environmental “noise”. That is, as you would expect, quite difficult to arrange. So far, the most advanced quantum computer in the world, IBM’s Osprey, has 433 qubits. This might not sound like much, but as the company points out “the number of classical bits that would be necessary to represent a state on the Osprey processor far exceeds the total number of atoms in the known universe”. What they don’t say is that it only works for about 70 to 80 millionths of a second before being overwhelmed by noise. Not only that, but the calculations it can make have very limited applications. As Kaku himself notes: “A workable quantum computer that can solve real-world problems is still many years in the future.” Some physicists, such as Mikhail Dyakonov at the University of Montpellier, believe the technical challenges mean the chances of a quantum computer “that could compete with your laptop” ever being built are pretty much zero.
Kaku brushes this off. He points to the billions of dollars being poured into quantum research – “the Gold Rush is on” he says – and the way intelligence agencies have been warning about the need to get quantum-ready. That’s hardly proof positive they’ll live up to expectations – it could be tulip mania rather than a gold rush. He shrugs: “Life’s a gamble.”
In any case, he’s far from the only true believer. Corporations such as IBM, Google, Microsoft and Intel are investing heavily in the technology, as is the Chinese government, which has developed a 113 qubit computer called Jiuzhang. So, assuming for a moment quantum dreams do become a reality: is it responsible to accentuate the positive, as Kaku does? What about the possibility of these immense capabilities being used for ill?
“Well, that’s the universal law of technology, that [it] can be used for good or evil. When humans discovered the bow and arrow, we could use that to bring down game and feed people in our tribe. But of course, the bow and arrow can also be used against our enemies.”
Advances in physics, in particular, have always raised the prospect of new and more fearsome weapons. But you can’t hold back research as a result: you make the discoveries, then you deal with the consequences. “That’s why we regulate nuclear weapons. Nuclear weapons are a rather simple consequence of Einstein’s E=mc2. And they have to be regulated, because the ‘E’ would be enough to destroy humanity on planet Earth. At some point, we’re going to reach the boundaries of this technology, where it impacts negatively on society. Right now, I can see a lot of benefits.”
In any case, for Kaku, knowledge is power. It’s part of the reason he’s moved from the lab to TV, radio and books. “The whole purpose of writing books for the public is so that [they] can make educated, reasonable, wise decisions about the future of technology. Once technology becomes so complicated that the average person cannot grasp it, then there’s big trouble, because then people with no moral compass will be in charge of the direction of that technology.”
There are other reasons, as well. From an early age, Kaku was, unsurprisingly, a science fiction nut. But he wasn’t content to simply swallow the stories, and wanted to know if they were really possible, whether the laws of physics might verify or contradict them. “And in the science section, there was nothing, absolutely nothing. And I was [also] fascinated by Einstein’s dream of a theory of everything, a unified field theory. Again I found nothing, not a single book, on Einstein’s great dream. And I said to myself, when I grow up, and I become a theoretical physicist, I want to write papers on this subject. But I also want to write for myself as a child, going to the library and being so frustrated that there was nothing for me to read. And that’s what I do.”
Kaku’s parents were among those American citizens of Japanese descent who were interned during the second world war, despite having been born in the country. Like his father, he was raised in Palo Alto, California, the “ground zero” of the tech revolution. The irony isn’t lost on him. “I saw Silicon Valley grow from nothing. When I was a child, it was all alfalfa fields, apple orchards. I used to play in the apple orchards of what is now Apple,” he chuckles. If his predictions about the quantum revolution are correct, it could soon be transformed again. “Silicon Valley could become a rust belt … a junkyard of chips that no one uses any more because they’re too primitive.” Or, more likely, a gleaming new centre of quantum computation, as today’s tech giants scramble to redeploy their immense intellectual and financial capital. Whether Kaku’s quantum revolution lives up to the hype remains to be seen. But if he is right and all that is digital passes into dust, we’re in for one hell of a ride.“