Tech
US-INTERNET-AI-META
Meta CEO Mark Zuckerberg presents the new Ray-Ban display glasses at the 2025 Meta Connect conference (Benjamin Legendre/Getty Images)

Watch: Meta virtual reality conference bloopers

Meta Connect was tough to watch.

Meta unveiled a host of seemingly innovative products and features at its annual virtual reality conference, Meta Connect, but what stood out were all the mess-ups.

Tech product events usually aren’t perfect, but they’re highly rehearsed and controlled environments, so they’re rarely this bad — especially for a company as big and as practiced as Meta.

While the prerecorded videos of the products in use were slick and highly produced, some of the live demos simply failed.

“Glasses are the ideal form factor for personal superintelligence because they let you stay present in the moment while getting access to all of these AI capabilities to make you smarter, help you communicate better, improve your memory, improve your senses, CEO Mark Zuckerberg reiterated at the start of the event, but the ensuing bloopers certainly didn’t make it feel that way.

In the very first demo of Live AI, Chef Jack Mancuso tried to get a recipe for steak sauce only to have the AI ignore his repeated requests and skip steps:

Perhaps the most painful example of the night was when Zuckerberg attempted to take a video call from CTO Andrew Boz Bosworth, who’s recently been spending time in his new role as an Army officer. They tried five times and eventually Boz had to come onstage.

“We’re going to have Boz come out here and we’re just going to go to the next thing that I wanted to show and hope that will work,” Zuckerberg said, visibly stressed.

Throughout the event, Zuckerberg fumbled words and blamed the Wi-Fi for the glitches. It was uncomfortable, strange, and distracted from the product lineup, which included new Ray-Bans with a built-in display that’s controlled by a wristband and Hyperscape Capture, tech that allows Quest users to quickly scan a room to create a virtual version of it.

“This isn’t a prototype,” Zuckerberg said of the Ray-Ban Display glasses with the wristband earlier in the event. “This is here. This is ready to go and you’re going to be able to buy them in a couple of weeks.”

But if the tech doesn’t work for Zuckerberg, who’s so practiced with the wristband he can type 30 words per minute, what are the rest of us who aren’t the CEO of the company to expect in real-life, day-to-day use cases?

Fortunately for Meta, analysts seem to have looked past the bloopers, with Bank of America, Morgan Stanley, JPMorgan, and Cantor Fitzgerald reiterating “buy” ratings after the event. Meta stock is up 0.5% premarket.

More Tech

See all Tech
Congress Considers Bill To Force Sale Of TikTok

By some measures, TikTok has grown bigger than Facebook or Instagram in the US

Oracle’s potential TikTok takeover is a big deal to US competitors like Meta and Snap.

tech

Nvidia, Microsoft, OpenAI, CoreWeave pledge $42 billion investment in UK AI projects during Trump’s visit

Nvidia, Microsoft, and CoreWeave announced pledges to invest tens of billions to build out the UK’s AI infrastructure.

Coinciding with President Trump’s visit to the UK, the companies announced new data centers, hundreds of thousands of Nvidia Blackwell GPUs, and support for the UK’s sovereign AI programs.

Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman are joining Trump for the visit.

Nvidia, CoreWeave, and UK AI infrastructure startup Nscale announced plans to roll out 120,000 Blackwell GPUs in UK data centers, including OpenAI’s “Stargate UK” data center project.

Part of the UK’s sovereign AI initiatives include the development of the country’s own “UK-LLM” and “Isambard-AI,” the UK’s most powerful supercomputer, which uses Nvidia’s Blackwell GPUs.

Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman are joining Trump for the visit.

Nvidia, CoreWeave, and UK AI infrastructure startup Nscale announced plans to roll out 120,000 Blackwell GPUs in UK data centers, including OpenAI’s “Stargate UK” data center project.

Part of the UK’s sovereign AI initiatives include the development of the country’s own “UK-LLM” and “Isambard-AI,” the UK’s most powerful supercomputer, which uses Nvidia’s Blackwell GPUs.

tech

Amazon launches AI chatbot to help create and distribute ads and ad agency investors don’t care

Amazon has launched a “creative partner” AI chatbot to help small businesses create ads and distribute them. The tool, currently in beta, helps users create the ads themselves, including video, with text prompts and then can place them across Amazon’s ad inventory, including outside websites and platforms Amazon has deals with, including Netflix.

Typically an announcement like this one pummels big advertising firms, whose livelihoods may or may not be threatened by the tech, but today Omnicom, Interpublic, WPP aren’t sinking on the news.

But perhaps the continuous stream of AI ad tool announcements from tech giants like Google, Amazon, and Meta is already baked into ad agencies’ stock prices. The ad agencies listed above are all down for the year.

Or perhaps these tools really are only for small businesses that can’t afford to work with big ad agencies.

“We’re not talking about professional marketers. These are customers that really need our help growing their business,” Jay Richman, Amazon’s vice president of product and technology, told The Wall Street Journal. Meta CEO Mark Zuckerberg, whose company expects to fully automate ad creation next year, said something similar on the company’s latest earnings call.

Typically an announcement like this one pummels big advertising firms, whose livelihoods may or may not be threatened by the tech, but today Omnicom, Interpublic, WPP aren’t sinking on the news.

But perhaps the continuous stream of AI ad tool announcements from tech giants like Google, Amazon, and Meta is already baked into ad agencies’ stock prices. The ad agencies listed above are all down for the year.

Or perhaps these tools really are only for small businesses that can’t afford to work with big ad agencies.

“We’re not talking about professional marketers. These are customers that really need our help growing their business,” Jay Richman, Amazon’s vice president of product and technology, told The Wall Street Journal. Meta CEO Mark Zuckerberg, whose company expects to fully automate ad creation next year, said something similar on the company’s latest earnings call.

tech

Report: Anthropic’s refusal to allow Claude to be used for surveillance irks White House

The Trump administration’s warm embrace of AI companies has led to many federal agencies using chatbots from OpenAI, Google, and Anthropic for many different applications.

Like its competitors, Anthropic is offering the government version of its chatbot — Claude for Government — for $1 per year to any agency that requests it, through the General Services Administration.

Semafor reports that contractors working for federal law enforcement agencies have encountered an obstacle: Anthropic’s policies don’t permit law enforcement to use Claude for surveillance applications. According to the report, Anthropic’s refusal to carve out an exception for federal law enforcement applications has “deepened hostility to the company” in the White House.

Under a section in Anthropic’s policy titled, “Do Not Use for Criminal Justice, Censorship, Surveillance, or Prohibited Law Enforcement Purposes,” the company explicitly prohibits the use of its products to “target or track a person’s physical location, emotional state, or communication without their consent, including using our products for facial recognition, battlefield management applications or predictive policing.”

Semafor reports that contractors working for federal law enforcement agencies have encountered an obstacle: Anthropic’s policies don’t permit law enforcement to use Claude for surveillance applications. According to the report, Anthropic’s refusal to carve out an exception for federal law enforcement applications has “deepened hostility to the company” in the White House.

Under a section in Anthropic’s policy titled, “Do Not Use for Criminal Justice, Censorship, Surveillance, or Prohibited Law Enforcement Purposes,” the company explicitly prohibits the use of its products to “target or track a person’s physical location, emotional state, or communication without their consent, including using our products for facial recognition, battlefield management applications or predictive policing.”

Latest Stories

Sherwood Media, LLC produces fresh and unique perspectives on topical financial news and is a fully owned subsidiary of Robinhood Markets, Inc., and any views expressed here do not necessarily reflect the views of any other Robinhood affiliate, including Robinhood Markets, Inc., Robinhood Financial LLC, Robinhood Securities, LLC, Robinhood Crypto, LLC, or Robinhood Money, LLC.