I just published a story that sets out some of the ways AI language models can be misused. I have some bad news: It’s stupidly easy, it requires no programming skills, and there are no known fixes. For example, for a type of attack called indirect prompt injection, all you need to do is hide a prompt in a cleverly crafted message on a website or in an email, in white text that (against a white background) is not visible to the human eye. Once you’ve done that, you can order the AI model to do what you want. ...
Allowing these language models to pull data from the internet gives hackers the ability to turn them into “a super-powerful engine for spam and phishing,” says Florian Tramèr, an assistant professor of computer science at ETH Zürich who works on computer security, privacy, and machine learning. ...
Unlike the spam and scam emails of today, where people have to be tricked into clicking on links, these new kinds of attacks will be invisible to the human eye and automated. ...
Deeper Learning
Chinese creators use Midjourney’s AI to generate retro urban “photography”
“Inside the Nimitz, which is housed in a massive, unfolding 18-wheeler, participants go through a video briefing before strapping on an Oculus Rift headset and a piece of wearable technology called a SubPac that percusses in real time to the sounds of the mission,” the Navy said in a release. “Visitors navigate the mission using a cutting-edge steering wheel and throttle system that replicate the actual sensation of piloting a high-speed, Navy Special Warfare Combatant Craft.”
At the conclusion of the mission, participants will move to a “debriefing station” where they’ll receive feedback and a performance grade.
An Agency Watchdog Rather Than a Legislative Cure.
As we have urged in the context of digital platforms, this is precisely the kind of situation that calls out for an expert administrative agency. The need for flexibility makes drafting legislation designed to consider all possible uses virtually impossible. ...
Amazon’s Kindle Unlimited young adult romance bestseller list was filled with dozens of AI-generated books of nonsense on Monday and Tuesday. As of Wednesday morning, Amazon appeared to have taken action against the books, but the episode shows that people are spamming AI-generated nonsense to the platform and are finding a way to monetize it. ...
“I honestly thought Amazon had a handle on the click farms. CLEARLY NOT.”
The Vatican is getting in on the AI craze. The Holy See has released a handbook on the ethics of artificial intelligence as defined by the Pope.
The guidelines are the result of a partnership between Francis and Santa Clara University’s Markkula Center for Applied Ethics. Together, they’ve formed a new organization called the Institute for Technology, Ethics, and Culture (ITEC). The ITEC’s first project is a handbook titled Ethics in the Age of Disruptive Technologies: An Operational Roadmap, meant to guide the tech industry through the murky waters of ethics in AI, machine learning, encryption, tracking, and more.
Google has killed a project internally known as "Iris" that it established to build augmented reality glasses, according to Business Insider. The tech giant worked on the glasses' development for several years, but it reportedly shelved Iris following waves of layoffs and company reshuffles over the past few months. Another event that factored into Google's decision was the departure of Clay Bavor, the company's former chief of augmented and virtual reality, to form a startup with Salesforce co-CEO Bret Taylor. In addition, Google reportedly kept changing its strategy for the Iris glasses, which became a source of frustration for team members working on the project. ...
While the company has stopped working on its own AR hardware, Business Insider says it still very much has big augmented reality ambitions. Instead of building its own hardware, Google has apparently chosen to focus on creating an "Android for AR" instead. ...
AI Pragmatists believe more strongly in embracing the ‘genius of the AND’, instead of succumbing to the ‘tyranny of the OR’. An AI Pragmatist understands that adoption of AI across all aspects of their life is inevitable, yet they want regulation and market pressures to ensure those systems are developed safely and responsibly. They believe that some uses of AI should be outright banned, most fully disclosed, and others can simply replace existing software experiences. They are excited by potential AI innovations to provide benefits in healthcare and education, but also want to ensure those advancements don’t leave people behind. Similarly, they know that biases exist in every dataset and model because of pre-existing human biases, yet they want to find ways to reduce the impact of those biases on AI outcomes. ...
RiverSpring Living, a services provider for more than 18,000 New Yorkers, is collaborating with VR company Rendever to help people who are being discharged from hospital or rehab to be reacquainted with their living spaces, the company announced Tuesday.
The VR platform allows older adults to explore their home and possibly review new additions such as hand railings before actually returning.
The tech could help prevent falls — an issue that affects 40% of older adults after being discharged — and reduce readmission rates, Rendever said. ...
The platform relies on synthesizing videos and pictures that are taken of the patients’ home. ...
Meta Quest users can subscribe to a new service that gives them access to the top two titles every month. Dubbed Meta Quest+, the virtual reality subscription costs $7.99 per month or $59.99 annually. ...
The first games being offered are the action-rhythm game FPS Pistol Whip and the arcade adventure game Pixel Ripped 1995. In August, players will get to experience the games Walkabout Mini Golf and Mothergunship: Forge.
Users keep all the titles as long as they remain subscribed to Meta Quest+. ...
