Weekly AI Recap: Ilya Sutskever launches AI company, Runway unveils Gen-3 Alpha

“SSI is our mission, our name, and our entire product roadmap, because it is our sole focus,” the company wrote in an X post on Wednesday. “Our team, investors, and business model are all aligned to achieve SSI.”

Superintelligence is within reach.

Building safe superintelligence (SSI) is the most important technical problem of our​​ time.

We've started the world’s first straight-shot SSI lab, with one goal and one product: a safe superintelligence.

It’s called Safe Superintelligence…

— SSI Inc. (@ssi) June 19, 2024

“Superintelligence,” a term coined by Oxford philosopher Nick Bostrom in a 2014 book by the same name, refers to an as-yet hypothetical AI that vastly exceeds human intelligence and can act autonomously.

Sutskever was a leader in the brief ousting of OpenAI CEO Sam Altman in November, which resulted in a shuffling of the company’s board and Altman’s reinstatement at the helm. He left the company last month, along with AI safety researcher Jan Leike, who quickly found a new role at Anthropic. (Sutskever and Leike had previously run OpenAI’s Superaligment team.)

Runway releases new video-generating model specializing in human likenesses

On Monday, AI company Runway introduced Gen-3 Alpha, a new AI model that “excels at generating expressive human characters with a wide range of actions, gestures, and emotions, unlocking new storytelling opportunities,” according to a company blog post.

It’s also capable of producing videos in a wide variety of artistic styles and can delve well into surrealist or fantastic territory. One sample video shared in the blog post, for example, was generated from the prompt: “View out a window of a giant strange creature walking in rundown city at night, one single street lamp dimly lighting the area.”

[Watch: Runway’s head of growth marketing Emily Golden in conversation with The Drum reporter Webb Wright at The Drum Live 2024.]

Audiences uncomfortable with AI-generated content in journalism, report finds

A significant percentage of Americans (52%) and Europeans (47%) are uncomfortable with news content created mostly by AI, according to this year’s Digital News Report from the Reuters Institute for the Study of Journalism.

Survey respondents reported feeling particularly wary of AI playing a role in journalism that focused on more sensitive subjects, like politics and crime.

The report – which surveyed close to 100,000 people across 47 countries – arrives at a time when a growing number of major news organizations are inking content deals with OpenAI (the maker of ChatGPT) and exploring other means of incorporating AI into their business models.

“Overall, we are still at the early stages of journalists’ usage of AI, but this also makes it a time of maximum risk for news organizations,” the Reuters Institute noted in its report. “Our data suggest that audiences are still deeply ambivalent about the use of the technology, which means that publishers need to be extremely cautious about where and how they deploy it.”

EU opens AI Office

On Sunday, the European Commission (EC) – the European Union’s executive arm – opened its AI Office, the latest phase in the ongoing implementation of the AI Act.

The Office is responsible for coordinating with the EU’s 27 member states to adopt the Act and for overseeing safety evaluations for “general purpose” AI models (a category that the Office has yet to concretely define).

Slated to go into effect later this summer, the AI Act will establish regulatory guardrails around approximately 15% of the AI-powered products and services that are currently operating in the EU. The remaining 85% will not fall under the purview of the Act “because they pose low to absolutely no risk in the EU,” EC spokesperson Thomas Regnier told The Drum.

McDonald’s remains committed to AI following AOT blunder

Fast food giant McDonald’s has pulled the plug on an AI-powered drive-thru assistant that has been deployed in more than 100 of its locations, Restaurant Business reported on June 14.

Dubbed Automated Order Taker (AOT), the system was developed through a partnership between McDonald’s and IBM.

Following its introduction in 2021, a number of customers posted videos to TikTok showing bizarre experiences with AOT, which appeared to occasionally misunderstand order requests and behave erratically.

Though McDonald’s says it will discontinue AOT at the end of this year, it remains keen to explore business opportunities with AI through its partnership with IBM. “As we move forward, our work with IBM has given us the confidence that a voice-ordering solution for drive-thru will be part of our restaurants’ future,” a company spokesperson told The Drum. “We see tremendous opportunity in advancing our restaurant technology and will continue to evaluate long-term, scalable solutions that will help us make an informed decision on a future voice ordering solution by the end of the year.”

For more on the latest happenings in AI, web3 and other cutting-edge technologies, sign up for The Emerging Tech Briefing newsletter.