Tag Archives: author_name|Mariella Moon

Two more OpenAI leaders are leaving the company

Posted on by

Two key OpenAI personnel are leaving the company, while another one is taking a months-long leave of absence. As reported by The Information, OpenAI president Greg Brockman is taking a sabbatical, while his fellow co-founder John Schulman is bidding the company farewell and joining its rival Anthropic. Peter Deng, its VP for consumer product who joined last year, is leaving OpenAI, as well. Deng used to lead product developments at Meta and Uber.

In a post on X/Twitter, Brockman said his leave of absence will last through end of year and that it's his "first time to relax" since the founding of OpenAI nine years ago. He also reportedly assured staff members that he's coming back after his vacation. Brockman temporarily left OpenAI last year when the company's board ousted its CEO, Sam Altman. They were both reinstated just a few days later, whereas the board was disbanded and replaced.

John Schulman also posted the note he sent to OpenAI staff on X, where he said that his departure stems from his "desire to deepen [his] focus on AI alignment" and to "return to hands-on technical work." He clarified that he wasn't leaving OpenAI due to lack of support for alignment research and that the company's executives have apparently been "very committed to investing in this area." At OpenAI, Schulman helped lead the post-training team that refined its large language models. After the company disbanded the "Superalignment" team that was in charge of steering powerful AI models that could pose a threat to humanity in the future, OpenAI said Schulman would be in charge of any future safety efforts. 

These departures follow several major previous changes in OpenAI's rank of leaders. Before the Superalignment team was disbanded, Jan Leike, one of its heads, wrote on X that "safety culture and processes have taken a backseat to shiny products" within the company over the past years. Company co-founder and Chief Scientist Ilya Sutskever also left in May. He was involved in Altman's dismissal as a board member, though he later said that he regretted his participation in the board's actions. While he remained Chief Scientist after the event, he reportedly never truly returned to his duties. The Information also notes that OpenAI recently hired its first CFO and chief product officer, which may have had an effect on Deng's role within the company. 

This article originally appeared on Engadget at https://www.engadget.com/ai/two-more-openai-leaders-are-leaving-the-company-110012003.html?src=rss

Prison Architect 2 is delayed indefinitely

Posted on by

Prison Architect 2, the sequel to the 2015 cult hit, isn't coming out this year as its creators had planned. The prison construction simulator's publisher, Paradox Interactive, has announced that its release has been delayed indefinitely. In its post, Paradox has admitted that its internal reviews and beta testing feedbacks surfaced the areas that need more attention, mainly the game's performance and content. "We need to raise the quality a bit more to meet the standards we'd like to achieve with this sequel," it wrote. 

This latest development comes after a series of delays. Prison Architect 2 was supposed to come out on March 26, but it was pushed back a couple of times so that its developers at Double Eleven and Kokku could resolve issues concerning memory usage and minimum spec configuration. This time, Paradox decided not to set a new release date. It says it will announce a new one once the Prison Architect 2 teams are done reworking the game's scope. 

Earlier this year, Paradox Interactive also had to deal with the overwhelmingly negative reception for Cities: Skylines 2's Beach Properties asset pack. Players weren't happy with the $10 DLC's contents, especially since they're still waiting for patches for the main game that was riddled with visual and mechanical bugs when it was released in 2023. Prison Architect 2's delay could prevent a repeat of what happened to Cities: Skylines.

Paradox explained that improving the game will take time, because it has "deep systems that all interact with each other." If one area is fixed, other parts of the game need to be fixed or adjusted, as well. For now, the publisher and the developers will refund all pre-orders. Those who purchased from Paradox directly will automatically be reimbursed, but those who purchased from the Xbox, PlayStation or Steam stores will have to go through those platforms to get their refund. Paradox will remove the option to pre-order the game entirely and will add the supposed pre-order bonus to the base game instead.

This article originally appeared on Engadget at https://www.engadget.com/prison-architect-2-is-delayed-indefinitely-160038125.html?src=rss

Meta is reportedly offering millions to use Hollywood voices in AI projects

Posted on by

A future artificial intelligence product by Meta could have you chatting with celebrities. According to Bloomberg and The New York Times, the company is in talks with Awkwafina, Judi Dench and Keegan-Michael Key, among other celebrities from various Hollywood agencies for its AI projects. The company apparently intends to incorporate their voices into a conversational generative AI-slash-digital assistant called MetaAI, which is similar to Siri and Google Assistant. 

Meta plans to record their voices and to secure the right to use them for as many situations as possible across Facebook, Messenger, Instagram, WhatsApp and even the Ray-Ban Meta glasses. Bloomberg says negotiations have started and stopped many times, because both sides can't seem to agree with the terms for use. For now, they seemed to have settled on a time limit, meaning any voice the company records can only be used over a set period. However, the deals with the actors could be renewed or extended by the time their contract is up. 

The actors' representatives are still looking to negotiate for stricter limits, though SAG-AFTRA has reportedly reached an agreement with Meta on terms. SAG-AFTRA, if you'll recall, fought for the establishment of provisions to protect actors from the threat of job loss due to AI when it went on strike last year. Under those terms, a company will have to pay actors and obtain their consent before it can use their AI-generated likeness. If Meta reaches a deal with the actors it's talking to, it could pay them millions of dollars in fees. 

Meta is looking to finalize deals before its Connect conference in September, The Times says, where it's expected to launch a bunch of AI products. During the same event last year, the company also introduced a chatbot platform with 28 "characters" voiced by celebrities, including Snoop Dogg, Paris Hilton, Dwyane Wade and Kendall Jenner. The Information reports that Meta has just quietly scrapped that project, and the celebrity chatbots' pages on Facebook and Instagram are no longer available. 

This article originally appeared on Engadget at https://www.engadget.com/meta-is-reportedly-offering-millions-to-use-hollywood-voices-in-ai-projects-121019385.html?src=rss

Nintendo profits fall 55 percent as people save their cash for the Switch 2

Posted on by

People are so excited for the next-gen Switch, they're likely holding off on buying Nintendo's current consoles and games. At least that's what the company's latest earnings report seems to indicate. For the quarter ending on June 30, Nintendo posted a net profit of 80.9 billion Japanese Yen, which is higher than its forecast but over 50 percent lower than its net profit for the same period last fiscal year. In addition, the company said it only sold 2.1 million Switch consoles for the quarter. That means it experienced a 46.3 percent decline on unit sales year-on-year. Even its games didn't sell well, seeing as Nintendo posted a software sales figure that's 41.3 percent lower than last fiscal year's at 30.64 million units sold. 

In its report, Nintendo admits that the low sales figures for games was caused by the lack of big releases, such as the previous year's The Legend of Zelda: Tears of the Kingdom. The Super Mario Bros. Movie also helped "energize" its business back then. But since hardware sales for this quarter are similar to the previous one's, Nintendo considers its Switch sales to be stable. 

Nintendo is expected to launch its "Switch 2" console soon. It was expected to come out sometime this year, but according to reports published in the previous months, it will be released in early 2025 instead. There's still very little known about the upcoming console, but rumors say it will have backwards compatibility, as well as 4K capabilities. 

This article originally appeared on Engadget at https://www.engadget.com/nintendo-profits-fall-55-percent-as-people-save-their-cash-for-the-switch-2-140019403.html?src=rss

Intel makes good on CPU instability issues by extending warranties by two years

Posted on by

Intel is extending the warranties for its controversial Core 13th and 14th-gen processors by two years, it has announced in a community post. The company says it will share more details in coming days, but for now, customers just waiting for their computers to conk out can at least know that they may not have to spend money to replace their processors. Intel revealed in July that after extensive analysis, it found that elevated operating voltage was causing these particular processor models' instability issues for a lot of people.

A microcode algorithm has apparently been sending incorrect voltage requests to the processor, causing users' computers to crash. The company is working on a patch that it plans to release in mid-August, but for some people, it may be too late: As Tom's Hardware notes, the patch will not fix processors that are already crashing. An indie gaming studio called Alderon Games reported that based on its personnel's observations, the processors' failure rate is 100 percent. Even CPUs that work well deteriorate and fail in the end. That's why an extended warranty is very much welcome, especially since some models only have a year-long warranty. 

"Intel is committed to making sure all customers who have or are currently experiencing instability symptoms on their 13th and/or 14th Gen desktop processors are supported in the exchange process," the company wrote in its announcement. It also admitted that "this has been a challenging issue to unravel and definitively root cause." For now, Intel advises those who purchased systems from computer manufacturers to reach out to the brand's support team. Meanwhile, people who purchased boxed CPUs for their PCs can contact Intel's customer support.

This article originally appeared on Engadget at https://www.engadget.com/intel-makes-good-on-cpu-instability-issues-by-extending-warranties-by-two-years-130010567.html?src=rss

Court blocks the FCC’s efforts to restore net neutrality… again

Posted on by

The Federal Communications Commission's voted to restore net neutrality protections back in April, but the process isn't as smooth-sailing as its proponents would like. According to Reuters and Fast Company, the Sixth Circuit US Court of Appeals has temporarily blocked the rules from taking effect because the broadband providers' legal case challenging their reinstatement will likely succeed. A group of cable, telecom and mobile internet providers sued the FCC shortly after its three Democrat commissioners voted to restore net neutrality protections

Under net neutrality rules, broadband services are classified as essential communications resources. That gives the FCC the power to regulate broadband internet and to prohibit providers from offering paid prioritization, which some ISPs have been using to charge bandwidth-heavy companies like Netflix additional fees. It will also prevent ISPs from blocking or slowing down traffic to specific websites. 

Net neutrality's opponents have long argued that the rules will put off investors. The group of providers that filed this recent case against the FCC said the rules' reinstatement would force them to "forego valuable new services, incur prohibitive compliance costs and pay more to obtain capital." In its decision, the court wrote that the "commission has failed to satisfy the high bar for imposing such regulations and that "net neutrality is likely a major question requiring clear congressional authorization."

The commission originally approved net neutrality rules back in 2015, though they have been in the works for years before that. Under the Trump administration, however, the FCC had voted to roll back the rules and to reclassify broadband internet services back to Title I, which means the agency would have less oversight on the industry. The rules were supposed to take effect on July 22 after the FCC voted to reinstate them, but a court blocked them from taking effect until August 5. Now, net neutrality's proponents will have to wait even longer. The appeals court has scheduled oral arguments discussing the issue for late October or early November, before or during the 2024 US presidential election. 

This article originally appeared on Engadget at https://www.engadget.com/court-blocks-the-fccs-efforts-to-restore-net-neutrality-again-123029311.html?src=rss

Boeing eats another $125 million loss over Starliner woes

Posted on by

Boeing has revealed that it has taken another $125 million in losses as a result of its Starliner spacecraft's delayed return from the ISS. As SpaceNews reports, the company has revealed the losses in a filing with the US Securities and Exchange Commission, along with more details about its earnings for the second quarter of the year. The company already posted $288 million in losses "primarily as a result of delaying" the Crew Flight Test mission in 2023. 

The first crewed Starliner flight took off in June with NASA astronauts Butch Wilmore and Sunita Williams on board. Boeing's spacecraft was only supposed to stay docked to the ISS for eight days before ferrying the astronauts back to Earth, but issues with its hardware prevented the mission from sticking to its original timeline. 

The company had to examine and find what caused the Starliner's degraded maneuvering thrusters while it was approaching the ISS. In addition, the helium leak that caused several delays to the spacecraft's launch seemed to have worsened, as well. Since June, the company has been putting the spacecraft through a series of tests. Just a few days ago, on July 27, it completed a hot fire test of the Starliner's reaction control system jets and made sure that the vehicle's helium leak rates remain within the acceptable margin. The tests were conducted with Williams and Wilmore onboard, because they're part of the preparations for the spacecraft's flight back home. 

NASA said the tests' results are still being reviewed. But once Boeing and the agency ensure that the Starliner is ready, they will set a date for the Starliner and the astronauts' return flight. 

This article originally appeared on Engadget at https://www.engadget.com/boeing-eats-another-125-million-loss-over-starliner-woes-130027376.html?src=rss

OpenAI vows to provide the US government early access to its next AI model

Posted on by

OpenAI will give the US AI Safety Institute early access to its next model as part of its safety efforts, Sam Altman has revealed in a tweet. Apparently, the company has been working with the consortium "to push forward the science of AI evaluations." The National Institute of Standards and Technology (NIST) has formally established the Artificial Intelligence Safety Institute earlier this year, though Vice President Kamala Harris announced it back in 2023 at the UK AI Safety Summit. Based on the NIST's description of the consortium, it's meant "to develop science-based and empirically backed guidelines and standards for AI measurement and policy, laying the foundation for AI safety across the world."

The company, along with DeepMind, similarly pledged to share AI models with the UK government last year. As TechCrunch notes, there have been growing concerns that OpenAI is making safety less of a priority as it seeks to develop more powerful AI models. There were speculations that the board decided to kick Sam Altman out of the company — he was very quickly reinstated — due to safety and security concerns. However, the company told staff members in an internal memo back then, that it was because of "a breakdown in communication."

In May this year, OpenAI admitted that it disbanded the Superalignment team it created to ensure that humanity remains safe as the company advances its work on generative artificial intelligence. Before that, OpenAI co-founder and Chief Scientist Ilya Sutskever, who was one of the team's leaders, left the company. Jan Leike, who was also one of the team's leaders, quit, as well. He said in a series of tweets that he had been disagreeing with OpenAI's leadership about the core priorities of the company for quite some time and that "safety culture and processes have taken a backseat to shiny products." OpenAI created a new safety group by the end of May, but it's led by board members that include Altman, prompting concerns about self-policing. 

This article originally appeared on Engadget at https://www.engadget.com/openai-vows-to-provide-the-us-government-early-access-to-its-next-ai-model-110017697.html?src=rss

Google makes it easier to remove explicit deepfakes from its search results

Posted on by

Google has rolled out updates for Search with the intention of making explicit deepfakes as hard to find as possible. As part of its long-standing and ongoing fight against realistic-looking manipulated images, the company is making it easier for people to get non-consensual fake imagery that features them removed from Search. 

It has long been possible for users to request for the removal of those kinds of images under Google's policies. Now, whenever it grants someone's removal request, Google will also filter all explicit results on similar searches about them. The company's systems will scan for any duplicates of the offending image and remove them, as well. This update could help alleviate some of the victim's fears if they're worried about the same image popping up again on other websites. 

In addition, Google has updated its ranking systems so that if a user specifically searches for explicit deepfakes with a person's name, the results will surface "high-quality, non-explicit content" instead. If there are news articles about that person, for instance, then the results will feature those. Based on Google's announcement, it seems it also has plans to school the user looking for deepfakes by showing them results that discuss their impact on society. 

Google doesn't want to wipe out results for legitimate content, like an actor's nude scene, in its bid to banish deepfakes from its results page, though. It admits it still has a lot of work to do when it comes to separating legitimate from fake explicit images. While that's still a work in progress, one of the solutions it has implemented is to demote sites that have received a high volume of removals for manipulated images in Search. That's "a pretty strong signal that it's not a high-quality site," Google explains, adding that the approach has worked well for other types of harmful content in the past.

This article originally appeared on Engadget at https://www.engadget.com/google-makes-it-easier-to-remove-explicit-deepfakes-from-its-search-results-130058499.html?src=rss

Meta explains why its AI claimed Trump’s assassination attempt didn’t happen

Posted on by

Meta has explained why its AI chatbot didn't want to respond to inquiries about the assassination attempt on Trump and then, in some cases, denied that the event took place. The company said it programmed Meta AI to not answer questions about an event right after it happens, because there's typically "an enormous amount of confusion, conflicting information, or outright conspiracy theories in the public domain." As for why Meta AI eventually started asserting that the attempt didn't happen "in a small number of cases," it was apparently due to hallucinations. 

An AI "hallucinates" when it generates false or misleading responses to questions that require factual replies due to various factors like inaccurate training data and AI models struggling to parse multiple sources of information. Meta says it has updated its AI's responses and admits that it should have done so sooner. It's still working to address its hallucination issue, though, so its chatbot could still be telling people that there was no attempt on the former president's life. 

In addition, Meta has also explained why its social media platforms had been incorrectly applying the fact check label to the photo of Trump with his fist in the air taken right after the assassination attempt. A doctored version of that image made it look like his Secret Service agents were smiling, and the company applied a fact check label to it. Because the original and doctored photos were almost identical, Meta's systems applied the label to the real image, as well. The company has since corrected the mistake. 

Trump's supporters have been crying foul over Meta AI's actions and have been accusing the company of suppressing the story. Google had to issue a response of its own after Elon Musk claimed that the company's search engine imposed a "search ban" on the former president. Musk shared an image that showed Google's autocomplete suggesting "president donald duck" when someone types in "president donald." Google explained that it was due to a bug affecting its autocomplete feature and said that users can search for whatever they want anytime. 

This article originally appeared on Engadget at https://www.engadget.com/meta-explains-why-its-ai-claimed-trumps-assassination-attempt-didnt-happen-120002196.html?src=rss