AI is driving massive demand for edge computing infrastructure, as industrial and commercial users need to process more data locally to take advantage of AI's capabilities, Ryan reports.

Why it matters: The shift follows years of big tech companies pushing organizations to migrate data to cloud computing services in remote data centers.

• Hardware and software providers alike are concluding that edge computing — moving processing power closer to where data is being generated — provides a bridge between 5G networks and cloud data centers. They're starting to package all three services.

Driving the news: Intel, AWS, Nokia and Ericsson announced collaborations on Feb. 12 to deliver edge AI services for manufacturing plants, transportation hubs such as ports and other complicated sites — such as the ancient Chichen Itza temple in Mexico.

• Labs are popping up — including one in St. Louis announced Feb. 15 by Intel, World Wide Technology and Federated Wireless — to help companies test and build customized private networks.
• "The whole industry is figuring out how to trim these [AI] models to fit at the edge without loss of accuracy," Sameer Vuyyuru, head of worldwide telecommunications business development at AWS, tells Axios.
• Many of the world's biggest tech companies are preparing to launch new edge AI hardware and software at Mobile World Congress in Barcelona, Spain, starting Feb. 25.

Be smart: The growth of edge AI is evidence that bigger is not always better in AI.

• Edge computing can enable faster data processing for time-sensitive applications and compliance with high security and privacy standards — suiting it to sectors such as health care and finance.

What's happening: "75% of data compute is moving to the edge," Kirk Skaugen, head of Lenovo's infrastructure business, tells Axios.

• "We're working with a large car company to get rid of 250,000 embedded PCs and replace them with probably 25,000 edge servers," he says.
• All the edge computing providers Axios spoke with said they're seeing high demand from organizations that operate in remote locations or have special security needs — from financial services to hospitals.

Details: Many new edge services are driven by customer demand.

• In the case of remote oil and gas rigs, owners found they were collecting massive amounts of data that became obsolete within seconds, so they needed ways to process that data quicker, Intel vice president Caroline Chan tells Axios.
• Lenovo sees a huge market for edge AI to support smart city technologies — from computer vision that responds to fires to helping vision-impaired people navigate streets.

By the numbers: The global market for edge computing is already worth over $200 billion a year.

Yes, but: To make better use of all the data they are producing, many large organizations will need to look for combinations of cloud computing, edge computing and private 5G networks to ensure their AI-enabled services run seamlessly.

OpenAI caused a stir on Thursday with its unveiling of Sora, its first tool that can turn a text prompt into a video of up to one minute in length.

Why it matters: While others, including Meta, Google and Runway, have their own text-to-video engines, the realism shown in sample videos elicited a powerful range of emotions.

Details: Sora is a diffusion model that is able to "generate complex scenes with multiple characters, specific types of motion, and accurate details of the subject and background," per OpenAI.

• Sora will be able to understand the nuances of the prompt as well as how various objects behave in the physical world.
• Sora also generates an entire video at once, rather than creating it frame by frame. That helps avoid what has been a continuity challenge with other approaches — ensuring a subject stays the same even when it goes out of view temporarily.

Between the lines: An OpenAI spokesperson stressed that it doesn't plan to make Sora broadly available any time soon as it continues to work on a range of safety issues, including efforts to reduce misinformation, hateful content and bias — and also to clearly label the output as generated by AI.

• OpenAI said in its blog post that it is making Sora available to some early testers, including both "red teamers" who will look to poke at potential security concerns, as well as a number of visual artists, designers and filmmakers, in order to get feedback on how the model could help them with their work.
• OpenAI got a range of feedback both based on the announcement itself, as well as a move by OpenAI CEO Sam Altman to create Sora designs based on a handful of prompts suggested on X (formerly Twitter).

What they're saying: Reactions to Sora represented a Rorschach test for how people already view the impact of generative AI, with a mix of excitement, awe, fear and revulsion.

• AI creator and ex-Googler Bilawal Sidhu said: "OpenAI just shattered the 'Visual' Turing Test, 'Is it real or is it fake?' isn't just for photos now. Videos too have fallen, and I couldn't be more excited."
• Filmmaker Justine Bateman, who has been highly critical of the technology, posted: "Every nanosecond of this #AI garbage is trained on stolen work by real artists. Repulsive."
• YouTube creator and tech reviewer Marques Brownlee was more measured, but noted: "If this doesn't concern you at least a little bit, nothing will."
• Meanwhile, Cristóbal Valenzuela, CEO of rival text-to-video service Runway, declared simply: "Game on."

New York Gov. Kathy Hochul is proposing legislation that would criminalize some deceptive and abusive uses of AI and require disclosure of AI in election campaign materials, her office tells Ryan.

Why it matters: New York is the country's second biggest AI center after California, and the proposals add to 65 draft AI bills already under consideration in the state.

• Both California and New York are using legislation and executive action to provide economic incentives for AI research while instituting new regulations to limit AI's harms.

Details: Hochul's proposed laws include establishing the crime of "unlawful dissemination or publication of a fabricated photographic, videographic, or audio record."

• It would make unauthorized uses of a person's voice "in connection with advertising or trade" a misdemeanor offense with up to a one year jail term.
• It would also expand New York's penal law to include unauthorized uses of artificial intelligence in coercion, criminal impersonation and identity theft.

Zoom out: Hochul's move is part of a wave of state-based AI legislation — now arriving at a rate of 50 bills per week — and often proposing criminal penalties for AI misuse.

What's next: Hochul's proposals are part of her executive budget proposal, meaning we'll know which parts are approved by New York's legislature by the end of March.

• Google launched Gemini Pro 1.5, an updated version of its model that offers improved performance as well as the ability to handle prompts with much more data, such as an hour of video, hours of audio or hundreds of thousands of words. (The Verge)
• A European Court found that the installation of backdoors for law enforcement into encrypted systems violates human rights. (Ars Technica)
• SAP announced a new AI unit and named Philipp Herzig as its chief AI officer.

It appears that a stingray at a North Carolina aquarium was either impregnated by a shark or managed to reproduce on her own.