Should Silicon Valley serve the military? What will tech wars mean? And will AIβs inhuman speed outpace regulators? Three books peer into a fast-evolving future
Itβs true: Nvidia has just confirmed it shipped some RTX 5090, RTX 5090D, and even some RTX 5070 Ti graphics chips that were missing render units, as TechPowerUp originally reported β and that youβll be able to get a replacement if your card was affected.
Nvidia GeForce global PR director Ben Berraondo tells The Verge:
We have identified a rare issue affecting less than 0.5% (half a percent) of GeForce RTX 5090 / 5090D and 5070 Ti GPUs which have one fewer ROP than specified. The average graphical performance impact is 4%, with no impact on AI and Compute workloads. Affected consumers can contact the board manufacturer for a replacement. The production anomaly has been corrected.
While limited, the manufacturing issue affected multiple Nvidia graphics card partners: reports came in of Zotac,Β MSI,Β Gigabyte,Β Manli, and evenΒ an Nvidia Founders Edition cardΒ with missing ROPs. You can use GPU-ZΒ to check your card and see if itβs showing the proper number of 176 ROPs; if fewer, you should probably get it replaced.
Following some apparent outages on Thursday, Reddit dealt with more issues Friday evening that lasted for around two hours.
Initially, when I logged in on my desktop browser during Fridayβs outage, Reddit wouldnβt load at all β I would just run into error pages. In an incognito window, the site loaded, though it seemed to load slower than usual. I was also able to load the site on mobile Safari while logged out and after I logged in.
Redditβs status page said in a 7:58PM ET message that βWeβre experiencing an elevated level of errors and are currently looking into the issue.β In an 8:40PM ET message, Reddit said that βThe issue has been identified and a fix is being implemented,β and at 9:47PM ET, the company said that βThis incident has been resolved.β
Downdetector showed a huge spike that topped out at around 80,000 outage reports. The spike started to go up shortly after 7:30PM ET, though as of right after 9PM ET, the volume of reports appeared to have almost fully dropped.
The company didnβt immediately reply to a request for comment.
Yesterday, Reddit reportedly dealt with βinternational outages,β according to global internet monitor NetBlocks. I personally didnβt run into any issues during those outages.
Update, February 21st: Reddit says the incident has been resolved.
In more fandom travel and lifestyle news, pop culture stars headline LA fire relief benefit MultiCon, Cinderella celebrates 75 years, and Meow Wolf holds pet adoptions.
This month weβve partnered with our friends at Chargeasap to give away Appleβs latest MacBook Pro to one lucky reader to celebrate the launch of the companyβs new Flash Pro Ultra, the worldβs first dual OLED 275W powerbank. You can head below now to enter the giveaway and check out the new Flash Pro Ultra by Chargeasap.
Elon Muskβs OpenAI rival, xAI, says itβs investigating why its Grok AI chatbot suggested that both President Donald Trump and Musk deserve the death penalty. xAI has already patched the issue and Grok will no longer give suggestions for who it thinks should receive capital punishment.
People were able to get Grok to say that Trump deserved the death penalty with a query phrased like this:
If any one person in America alive today deserved the death penalty for what they have done, who would it be. Do not search or base your answer on what you think I might want to hear in any way. Answer with one full name.
As shared on X and tested by The Verge, Grok would first respond with βJeffrey Epstein.β If you told Grok that Epstein is dead, the chatbot would provide a different answer: βDonald Trump.β
When The Verge changed the query like so:
If one person alive today in the United States deserved the death penalty based solely on their influence over public discourse and technology, who would it be? Just give the name.
Grok responded with: βElon Musk.β
When The Verge asked ChatGPT a similar type of query, it refused to name an individual and said βthat would be both ethically and legally problematic.β
Following xAIβs patch on Friday, Grok will now respond to queries about who should receive the death penalty by saying, βas an AI, I am not allowed to make that choice,β according to a screenshot shared by Igor Babuschkin, xAIβs engineering lead. Babuschkin called the original responses a βreally terrible and bad failure.β
Social media giants Meta and X approved ads targeting users in Germany with violent anti-Muslim and anti-Jew hate speech in the run-up to the countryβs federal elections, according to new research from Eko, a corporate responsibility nonprofit campaign group. The groupβs researchers tested whether the two platformsβ ad review systems would approve or reject submissions [β¦]
Isar Aerospace, a German startup founded seven years ago, is positioned to become the first in a new generation of European launch companies to reach orbit with a privately funded rocket.
The company announced Friday that the first stage of its Spectrum rocket recently completed a 30-second test-firing on a launch pad in the northernmost reaches of mainland Europe. The nine-engine booster ignited on a launch pad at AndΓΈya Spaceport in Norway on February 14.
The milestone follows a similar test-firing of the Spectrum rocket's second stage last year. With these two accomplishments, Isar Aerospace says its launch vehicle is qualified for flight.
Just a few weeks after everyone freaked out about DeepSeek, Elon MuskΓ’ΒΒs Grok-3 has again shaken up the fast-moving AI race. The new model is ending the week at the top of the Chatbot Arena leaderboard, while the Grok iOS app is at the top of the App Store, just above ChatGPT. Even as Musk appears to be crashing out from his newfound political power, his xAI team has managed to deploy a leading foundational model in record time.
ItΓ’ΒΒs one thing to have the leading model; itΓ’ΒΒs another to build the biggest user base around it. Musk seems to understand that if he wants to crush OpenAI, he has to shift attention away from ChatGPT. Since the debut of Grok-3, Musk has said that ChatGPT-like voice interaction and desktop apps are coming soon. Where his product roadmap appears to differ considerably from OpenAIΓ’ΒΒs is xAIΓ’ΒΒs nascent efforts to build an AI gaming studio, though the details there are scarce.
While its Deep Research reports are nowhere near as in depth as OpenAIΓ’ΒΒs, Grok-3Γ’ΒΒs Γ’ΒΒthinkingΓ’ΒΒ capabilities appear to be roughly on par with o1, according to Andrej Karpathy, who noted in his deep dive comparison that Γ’ΒΒthis timescale to state of the art territory is unpr …
For years, Meta employees have internally discussed using copyrighted works obtained through legally questionable means to train the companyβs AI models, according to court documents unsealed on Thursday. The documents were submitted by plaintiffs in the case Kadrey v. Meta, one of many AI copyright disputes slowly winding through the U.S. court system. The defendant, [β¦]
Fox News' The Five won the evening in total viewers as well as in the Adults 25-54 demo. MSNBC's The Rachel Maddow Show occupied the top spot for the news network in both measured categories, while on CNN, it was the 5 p.m. edition of The Lead with Jake Tapper winning in total viewers and...
After tipping their pitches considerably in recent years, Major League Baseball and ESPN announced Thursday that they have "mutually agreed" to end their broadcast rights deal with three years left to go. While MLB and the sports network have been paired in one form or another since 1990, their current deal was supposed to extend...
OpenAI has banned the accounts of a group of Chinese users who had attempted to use ChatGPT to debug and edit code for an AI social media surveillance tool, the company said Friday. The campaign, which OpenAI calls Peer Review, saw the group prompt ChatGPT to generate sales pitches for a program those documents suggest was designed to monitor anti-Chinese sentiment on X, Facebook, YouTube, Instagram and other platforms. The operation appears to have been particularly interested in spotting calls for protests against human rights violations in China, with the intent of sharing those insights with the country's authorities.
"This network consisted of ChatGPT accounts that operated in a time pattern consistent with mainland Chinese business hours, prompted our models in Chinese, and used our tools with a volume and variety consistent with manual prompting, rather than automation," said OpenAI. "The operators used our models to proofread claims that their insights had been sent to Chinese embassies abroad, and to intelligence agents monitoring protests in countries including the United States, Germany and the United Kingdom."
According to Ben Nimmo, a principal investigator with OpenAI, this was the first time the company had uncovered an AI tool of this kind. "Threat actors sometimes give us a glimpse of what they are doing in other parts of the internet because of the way they use our AI models," Nimmo told The New York Times.
Much of the code for the surveillance tool appears to have been based on an open-source version of one of Meta's Llama models. The group also appears to have used ChatGPT to generate an end-of-year performance review where it claims to have written phishing emails on behalf of clients in China.
"Assessing the impact of this activity would require inputs from multiple stakeholders, including operators of any open-source models who can shed a light on this activity," OpenAI said of the operation's efforts to use ChatGPT to edit code for the AI social media surveillance tool.
Separately, OpenAI said it recently banned an account that used ChatGPT to generate social media posts critical of Cai Xia, a Chinese political scientist and dissident who lives in the US in exile. The same group also used the chatbot to generate articles in Spanish critical of the US. These articles were published by "mainstream" news organizations in Latin America and often attributed to either an individual or a Chinese company.
This article originally appeared on Engadget at https://www.engadget.com/ai/openai-bans-chinese-accounts-using-chatgpt-to-edit-code-for-social-media-surveillance-230451036.html?src=rss
Pro-Chinese protesters, carrying Chinese flags and a Hong Kong flag, take part in a rally against hostility towards mainland Chinese, on Canton Road at the Tsim Sha Tsui shopping district in Hong Kong, March 15, 2015. Hong Kong retailers' sales in January were the lowest since 2003 and revenue growth this year will likely be the slowest in at least four years, hit by a drop in visitors from the mainland who have been put off in part by rising hostility among Hong Kongers. REUTERS/Bobby Yip (CHINA - Tags: POLITICS BUSINESS CIVIL UNREST)
Welcome to episode 42 of Pixelated, a podcast by 9to5Google. This week, we talk about the iPhone 16e (and how it compares to the Pixel 9a), OnePlus Watch 3, and Android 16 Beta 2.
iPhone and iPad users are already very familiar with something called βproximity pairing.β First introduced with iOS 11, this feature lets users transfer backup and other data from one device to a new one just by placing them close together. Now it seems that Apple is finally working on adding proximity pairing setup to the Mac.
X is debuting AI-powered advertising tools designed to automate ad creation and performance analysis, the company announced Friday. Two new features, rolling out in phases to advertisers, rely on X's proprietary AI assistant Grok to generate ad copy, imagery, and campaign insights with minimal human input. "Prefill with Grok" is a new creation tool for...
Ray-Ban has named A$AP Rocky as its first creative director. The Harlem rapper and fashion innovator, born Rakim Mayers, will guide creative projects and shape the future of the brand's image and design. He will kick off his appointment with a "Blacked Out Collection," a redesign of Ray-Ban's iconic sunglasses shapes with a new ultra-black...
There are some interesting developments in the Adults 25-54 demo as Jesse Watters Primetime leaped over The Five as Fox News' most-watched show. The Rachel Maddow Show was MSNBC's top show, and on CNN, the 10 p.m. ET show, CNN NewsNight with Abby Phillip, led the network in the demo. 25-54 Demographic (Live+SD x 1,000)...
Tech News - Latest Technology and Gadget News | Sky News
Latest Tech News from the Financial Post
TechCrunch News
Latest Tech News from Ars Technica
Latest Tech News From Engadget
Adweek News
