Новини

Tracking down accurate information about Philadelphia’s elections on Twitter used to be easy. The account for the city commissioners who run elections, @phillyvotes, was the only one carrying a blue check mark, a sign of authenticity.

But ever since the social media platform overhauled its verification service last month, the check mark has disappeared. That’s made it harder to distinguish @phillyvotes from a list of random accounts not run by the elections office but with very similar names.

The election commission applied weeks ago for a gray check mark — Twitter’s new symbol to help users identify official government accounts – but has yet to hear back from the Twitter, commission spokesman Nick Custodio said. It’s unclear whether @phillyvotes is an eligible government account under Twitter’s new rules.

That’s troubling, Custodio said, because Pennsylvania has a primary election May 16 and the commission uses its account to share important information with voters in real time. If the account remains unverified, it will be easier to impersonate – and harder for voters to trust – heading into Election Day.

Impostor accounts on social media are among many concerns election security experts have heading into next year’s presidential election. Experts have warned that foreign adversaries or others may try to influence the election, either through online disinformation campaigns or by hacking into election infrastructure.

Election administrators across the country have struggled to figure out the best way to respond after Twitter owner Elon Musk threw the platform’s verification service into disarray, given that Twitter has been among their most effective tools for communicating with the public.

Some are taking other steps allowed by Twitter, such as buying check marks for their profiles or applying for a special label reserved for government entities, but success has been mixed. Election and security experts say the inconsistency of Twitter’s new verification system is a misinformation disaster waiting to happen.

“The lack of clear, at-a-glance verification on Twitter is a ticking time bomb for disinformation,” said Rachel Tobac, CEO of the cybersecurity company SocialProof Security. “That will confuse users – especially on important days like election days.”

The blue check marks that Twitter once doled out to notable celebrities, public figures, government entities and journalists began disappearing from the platform in April. To replace them, Musk told users that anyone could pay $8 a month for an individual blue check mark or $1,000 a month for a gold check mark as a “verified organization.”

The policy change quickly opened the door for pranksters to pose convincingly as celebrities, politicians and government entities, which could no longer be identified as authentic. While some impostor accounts were clear jokes, others created confusion.

Fake accounts posing as Chicago Mayor Lori Lightfoot, the city’s Department of Transportation and the Illinois Department of Transportation falsely claimed the city was closing one of its main thoroughfares to private traffic. The fake accounts used the same photos, biographical text and home page links as the real ones. Their posts amassed hundreds of thousands of views before being taken down.

Twitter’s new policy invites government agencies and certain affiliated organizations to apply to be labeled as official with a gray check. But at the state and local level, qualifying agencies are limited to “main executive office accounts and main agency accounts overseeing crisis response, public safety, law enforcement, and regulatory issues,” the policy says.

The rules do not mention agencies that run elections. So while the main Philadelphia city government account quickly received its gray check mark last month, the local election commission has not heard back.

Election offices in four of the country’s five most populous counties — Cook County in Illinois, Harris County in Texas, Maricopa County in Arizona and San Diego County — remain unverified, a Twitter search shows. Maricopa, which includes Phoenix, has been targeted repeatedly by election conspiracy theorists as the most populous and consequential county in one of the most closely divided political battleground states.

Some counties contacted by The Associated Press said they have minimal concerns about impersonation or plan to apply for a gray check later, but others said they already have applied and have not heard back from Twitter.

Even some state election offices are waiting for government labels. Among them is the office of Maine Secretary of State Shenna Bellows.

In an April 24 email to Bellows’ communications director reviewed by The Associated Press, a Twitter representative wrote that there was “nothing to do as we continue to manually process applications from around the world.” The representative added in a later email that Twitter stands “ready to swiftly enforce any impersonation, so please don’t hesitate to flag any problematic accounts.”

An email sent to Twitter’s press office and a company safety officer requesting comment was answered only with an autoreply of a poop emoji.

“Our job is to reinforce public confidence,” Bellows told the AP. “Even a minor setback, like no longer being able to ensure that our information on Twitter is verified, contributes to an environment that is less predictable and less safe.”

Some government accounts, including the one representing Pennsylvania’s second-largest county, have purchased blue checks because they were told it was required to continue advertising on the platform.

Allegheny County posts ads for elections and jobs on Twitter, so the blue check mark “was necessary,” said Amie Downs, the county’s communications director.

When anyone can buy verification and when government accounts are not consistently labeled, the check mark loses its meaning, Colorado Secretary of State Jena Griswold said.

Griswold’s office received a gray check mark to maintain trust with voters, but she told the AP she would not buy verification for her personal Twitter account because “it doesn’t carry the same weight” it once did.

Custodio, at the Philadelphia elections commission, said his office would not buy verification either, even if it gets denied a gray check.

“The blue or gold check mark just verifies you as a paid subscriber and does not verify identity,” he said.

Experts and advocates tracking election discourse on social media say Twitter’s changes do not just incentivize bad actors to run disinformation campaigns — they also make it harder for well-meaning users to know what’s safe to share.

“Because Twitter is dropping the ball on verification, the burden will fall on voters to double check that the information they are consuming and sharing is legitimate,” said Jill Greene, voting and elections manager for Common Cause Pennsylvania.

That dampens an aspect of Twitter that until now had been seen as one of its strengths – allowing community members to rally together to elevate authoritative information, said Mike Caulfield, a research scientist at the University of Washington’s Center for an Informed Public.

“The first rule of a good online community user interface is to ‘help the helpers.’ This is the opposite of that,” Caulfield said. “It takes a community of people who want to help boost good information, and robs them of the tools to make fast, accurate decisions.”
…

Billionaire Warren Buffett said artificial intelligence may change the world in all sorts of ways, but new technology won’t take away opportunities for investors, and he’s confident America will continue to prosper over time.

Buffett and his partner Charlie Munger are spending all day Saturday answering questions at Berkshire Hathaway’s annual meeting inside a packed Omaha arena.

“New things coming along doesn’t take away the opportunities. What gives you the opportunities is other people doing dumb things,” said Buffett, who had a chance to try out ChatGPT when his friend Bill Gates showed it to him a few months back.

Buffett reiterated his long-term optimism about the prospects for America even with the bitter political divisions today.

“The problem now is that partisanship has moved more towards tribalism, and in tribalism you don’t even hear the other side,” he said.

Both Buffett and Munger said the United States will benefit from having an open trading relationship with China, so both countries should be careful not to exacerbate the tensions between them because the stakes are too high for the world.

“Everything that increases the tension between these two countries is stupid, stupid, stupid,” Munger said. And whenever either country does something stupid, he said the other country should respond with incredible kindness.

The chance to listen to the two men answer all sorts of questions about business and life attracts people from all over the world to Omaha, Nebraska. Some of the shareholders feel a particular urgency to attend now because Buffett and Munger are both in their 90s.

“Charlie Munger is 99. I just wanted to see him in person. It’s on my bucket list,” said 40-year-old Sheraton Wu from Vancouver. “I have to attend while I can.”

“It’s a once in a lifetime opportunity,” said Chloe Lin, who traveled from Singapore to attend the meeting for the first time and learn from the two legendary investors.

One of the few concessions Buffett makes to his age is that he no longer tours the exhibit hall before the meeting. In years past, he would be mobbed by shareholders trying to snap a picture with him while a team of security officers worked to manage the crowd. Munger has used a wheelchair for several years, but both men are still sharp mentally.

But in a nod to the concerns about their age, Berkshire showed a series of clips of questions about succession from past meetings dating back to the first one they filmed in 1994. Two years ago, Buffett finally said that Greg Abel will eventually replace him as CEO although he has no plans to retire. Abel already oversees all of Berkshire’s noninsurance businesses.

Buffett assured shareholders that he has total confidence in Abel to lead Berkshire in the future, and he doesn’t have a second choice for the job because Abel is remarkable in his own right. But he said much of what Abel will have to do is just maintain Berkshire’s culture and keep making similar decisions.

“Greg understands capital allocation as well as I do. He will make these decisions on the same framework that I use,” Buffett said.

Abel followed that up by assuring the crowd that he knows how Buffett and Munger have handled things for nearly six decades and “I don’t really see that framework changing.”

Although not everyone at the meeting is a fan. Outside the arena, pilots from Berkshire’s NetJets protested over the lack of a new contract and pro-life groups carried signs declaring “Buffett’s billions kill millions” to object to his many charitable donations to abortion rights groups.

Berkshire Hathaway said Saturday morning that it made $35.5 billion, or $24,377 per Class A share, in the first quarter. That’s more than 6 times last year’s $5.58 billion, or $3,784 per share.

But Buffett has long cautioned that those bottom line figures can be misleading for Berkshire because the wide swings in the value of its investments — most of which it rarely sells — distort the profits. In this quarter, Berkshire sold only $1.7 billion of stocks while recording a $27.4 billion paper investment gain. Part of this year’s investment gains included a $2.4 billion boost related to Berkshire’s planned acquisition of the majority of the Pilot Travel Centers truck stop company’s shares in January.

Buffett says Berkshire’s operating earnings that exclude investments are a better measure of the company’s performance. By that measure, Berkshire’s operating earnings grew nearly 13% to $8.065 billion, up from $7.16 billion a year ago.

The three analysts surveyed by FactSet expected Berkshire to report operating earnings of $5,370.91 per Class A share.

Buffett came close to giving a formal outlook Saturday when he told shareholders that he expects Berkshire’s operating profits to grow this year even though the economy is slowing down and many of its businesses will sell less in 2023. He said Berkshire will profit from rising interest rates on its holdings, and the insurance market looks good this year.

This year’s first quarter was relatively quiet compared to a year ago when Buffett revealed that he had gone on a $51 billion spending spree at the start of last year, snapping up stocks like Occidental Petroleum, Chevron and HP. Buffett’s buying slowed through the rest of last year with the exception of a number of additional Occidental purchases.

At the end of this year’s first quarter, Berkshire held $130.6 billion cash, up from about $128.59 billion at the end of last year. But Berkshire did spend $4.4 billion during the quarter to repurchase its own shares.

Berkshire’s insurance unit, which includes Geico and a number of large reinsurers, recorded a $911 million operating profit, up from $167 million last year, driven by a rebound in Geico’s results. Geico benefitted from charging higher premiums and a reduction in advertising spending and claims.

But Berkshire’s BNSF railroad and its large utility unit did report lower profits. BNSF earned $1.25 billion, down from $1.37 billion, as the number of shipments it handled dropped 10% after it lost a big customer and imports slowed at the West Coast ports. The utility division added $416 million, down from last year’s $775 million.

Besides those major businesses, Berkshire owns an eclectic assortment of dozens of other businesses, including a number of retail and manufacturing firms such as See’s Candy and Precision Castparts.

Berkshire again faces pressure from activist investors urging the company to do more to catalog its climate change risks in a companywide report. Shareholders were expected to brush that measure and all the other shareholder proposals aside Saturday afternoon because Buffett and the board oppose them, and Buffett controls more than 30% of the vote.

But even as they resist detailing climate risks, a number of Berkshire’s subsidiaries are working to reduce their carbon emissions, including its railroad and utilities. The company’s Clayton Homes unit is showing off a new home design this year that will meet strict energy efficiency standards from the Department of Energy and come pre-equipped for solar power to be added later.
…

Google is planning to make its search engine more “visual, snackable, personal and human,” with a focus on serving young people globally, The Wall Street Journal reported on Saturday, citing documents.

The move comes as artificial intelligence (AI) applications such as ChatGPT are rapidly gaining in popularity, highlighting a technology that could upend the way businesses and society operate.

The tech giant will nudge its service further away from “10 blue links,” which is a traditional format of presenting search results and plans to incorporate more human voices as part of the shift, the report said.

At its annual I/O developer conference in the coming week, Google is expected to debut new features that allow users to carry out conversations with an AI program, a project code-named “Magi,” The Wall Street Journal added, citing people familiar with the matter.

Generative AI has become a buzzword this year, with applications capturing the public’s fancy and sparking a rush among companies to launch similar products they believe will change the nature of work.

Google, part of Alphabet Inc., did not immediately respond to Reuters’ request for comment.
…

When Greg Brockman, the president and co-founder of ChatGPT maker OpenAI, was recently extolling the capabilities of artificial intelligence, he turned to “Game of Thrones.”

Imagine, he said, if you could use AI to rewrite the ending of that not-so-popular finale. Maybe even put yourself into the show.

“That is what entertainment will look like,” said Brockman.

Not six months since the release of ChatGPT, generative artificial intelligence is already prompting widespread unease throughout Hollywood. Concern over chatbots writing or rewriting scripts is one of the leading reasons TV and film screenwriters took to picket lines earlier this week.

Though the Writers Guild of America is striking for better pay in an industry where streaming has upended many of the old rules, AI looms as rising anxiety.

“AI is terrifying,” said Danny Strong, the “Dopesick” and “Empire” creator. “Now, I’ve seen some of ChatGPT’s writing and as of now I’m not terrified because Chat is a terrible writer. But who knows? That could change.”

AI chatbots, screenwriters say, could potentially be used to spit out a rough first draft with a few simple prompts (“a heist movie set in Beijing”). Writers would then be hired, at a lower pay rate, to punch it up.

Screenplays could also be slyly generated in the style of known writers. What about a comedy in the voice of Nora Ephron? Or a gangster film that sounds like Mario Puzo? You won’t get anything close to “Casablanca” but the barest bones of a bad Liam Neeson thriller isn’t out of the question.

The WGA’s basic agreement defines a writer as a “person” and only a human’s work can be copyrighted. But even though no one’s about to see a “By AI” writers credit at the beginning a movie, there are myriad ways that regenerative AI could be used to craft outlines, fill in scenes and mockup drafts.

“We’re not totally against AI,” says Michael Winship, president of the WGA East and a news and documentary writer. “There are ways it can be useful. But too many people are using it against us and using it to create mediocrity. They’re also in violation of copyright. They’re also plagiarizing.”

The guild is seeking more safeguards on how AI can be applied to screenwriting. It says the studios are stonewalling on the issue. The Alliance of Motion Picture and Television Producers, which bargains on the behalf of production companies, has offered to annually meet with the guild to go over definitions around the fast-evolving technology.

“It’s something that requires a lot more discussion, which we’ve committed to doing,” the AMPTP said in an outline of its position released Thursday.

Experts say the struggle screenwriters are now facing with regenerative AI is just the beginning. The World Economic Forum this week released a report predicting that nearly a quarter of all jobs will be disrupted by AI over the next five years.

“It’s definitely a bellwether in the workers’ response to the potential impacts of artificial intelligence on their work,” says Sarah Myers West, managing director of the nonprofit AI Now Institute, which has lobbied the government to enact more regulation around AI. “It’s not lost on me that a lot of the most meaningful efforts in tech accountability have been a product of worker-led organizing.”

AI has already filtered into nearly every part of moviemaking. It’s been used to de-age actors, remove swear words from scenes in post-production, supply viewing recommendations on Netflix and posthumously bring back the voices of Anthony Bourdain and Andy Warhol.

The Screen Actors Guild, set to begin its own bargaining with the AMPTP this summer, has said it’s closely following the evolving legal landscape around AI.

“Human creators are the foundation of the creative industries, and we must ensure that they are respected and paid for their work,” the actors union said.

The implications for screenwriting are only just being explored. Actors Alan Alda and Mike Farrell recently reconvened to read through a new scene from “M(asterisk)A(asterisk)S(asterisk)H” written by ChatGPT. The results weren’t terrible, though they weren’t so funny, either.

“Why have a robot write a script and try to interpret human feelings when we already have studio executives who can do that?” deadpanned Alda.

Writers have long been among notoriously exploited talents in Hollywood. The films they write usually don’t get made. If they do, they’re often rewritten many times over. Raymond Chandler once wrote “the very nicest thing Hollywood can possibly think to say to a writer is that he is too good to be only a writer.”

Screenwriters are accustomed to being replaced. Now, they see a new, readily available and inexpensive competitor in AI — albeit one with a slightly less tenuous grasp of the human condition.

“Obviously, AI can’t do what writers and humans can do. But I don’t know that they believe that, necessarily,” says screenwriter Jonterri Gadson (“A Black Lady Sketchshow”). “There needs to be a human writer in charge and we’re not trying to be gig workers, just revising what AI does. We need to tell the stories.”

Dramatizing their plight as man vs. machine surely doesn’t hurt the WGA’s cause in public opinion. The writers are wrestling with the threat of AI just as concern widens over how hurriedly regenerative AI products have been thrust into society.

Geoffrey Hinton, an AI pioneer, recently left Google in order to speak freely about its potential dangers. “It’s hard to see how you can prevent the bad actors from using it for bad things,” Hinton told The New York Times.

“What’s especially scary about it is nobody, including a lot of the people who are involved with creating it, seem to be able to explain exactly what it’s capable of and how quickly it will be capable of more,” says actor-screenwriter Clark Gregg.

The writers find themselves in the awkward position of negotiating on a newborn technology with the potential for radical effect. Meanwhile, AI-crafted songs by “Fake Drake” or “Fake Eminem” continue to circulate online.

“They’re afraid that if the use of AI to do all this becomes normalized, then it becomes very hard to stop the train,” says James Grimmelmann, a professor of digital and information law at Cornell University. “The guild is in the position of trying to imagine lots of different possible futures.”

In the meantime, chanting demonstrators are hoisting signs with messages aimed at a digital foe. Seen on the picket lines: “ChatGPT doesn’t have childhood trauma”; “I heard AI refuses to take notes”; and “Wrote ChatGPT this.”
…