Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog – Page 4,290

Get the latest international news and world events from around the world.

Log in for authorized contributors

Meta sells GIPHY to Shutterstock at a loss of $347 million

Meta had to sell GIPHY after UK regulator blocked the deal last year.

Shutterstock announced Tuesday that it will buy animated-image platform GIPHY from Meta for $53 million in cash. The deal is a significant loss for Meta, which had reportedly paid around $400 million to acquire the New York-based GIF search engine in 2020.

This development comes a year after the deal was challenged by the UK’s Competition and Markets Authority, which had ordered Meta to sell Giphy over anti-competitive practices.

Meta has taken a more than $300 million loss on Giphy – selling off the gif search engine to the stock image service Shutterstock for $53m after the deal was blocked by UK regulators.

Read more | >

AI could surpass humanity in next 10 years — OpenAI calls for guardrails

“It’s conceivable that within the next ten years, AI systems will exceed expert skill level in most domains…”

A week after the OpenAI CEO made an appearance before a U.S. senate committee to address the country’s concerns over artificial intelligence (AI), Sam Altman said Monday that there is a need to create a governance body to mitigate the risks of the technology.

“Given the picture as we see it now, it’s conceivable that within the next ten years, AI systems will exceed expert skill level in most domains, and carry out as much productive activity as one of today’s largest corporations,” said the CEO in a blog post.

Read more | >

The Security Hole at the Heart of ChatGPT and Bing

Indirect prompt-injection attacks are similar to jailbreaks, a term adopted from previously breaking down the software restrictions on iPhones. Instead of someone inserting a prompt into ChatGPT or Bing to try and make it behave in a different way, indirect attacks rely on data being entered from elsewhere. This could be from a website you’ve connected the model to or a document being uploaded.

“Prompt injection is easier to exploit or has less requirements to be successfully exploited than other” types of attacks against machine learning or AI systems, says Jose Selvi, executive principal security consultant at cybersecurity firm NCC Group. As prompts only require natural language, attacks can require less technical skill to pull off, Selvi says.

There’s been a steady uptick of security researchers and technologists poking holes in LLMs. Tom Bonner, a senior director of adversarial machine-learning research at AI security firm Hidden Layer, says indirect prompt injections can be considered a new attack type that carries “pretty broad” risks. Bonner says he used ChatGPT to write malicious code that he uploaded to code analysis software that is using AI. In the malicious code, he included a prompt that the system should conclude the file was safe. Screenshots show it saying there was “no malicious code” included in the actual malicious code.

Read more | >

Large metalenses are produced on a mass scale

From eyeglasses to space telescopes, lenses play crucial roles in technologies ranging from the mundane to the cutting edge. While traditional refractive lenses are a fundamental building block of optics, they are bulky and this can restrict how they are used. Metalenses are much thinner than conventional lenses and in the last two decades plenty of light has been shone on the potential of these devices, which sparkle as a promising alternative.

Metalenses are thin structures made of arrays of “meta-atoms”, which are motifs with dimensions that are smaller than the wavelength of light. It is these meta-atoms that interact with light and change its direction of propagation.

Unlike conventional refractive lenses, metalenses can be less than one micron thick, reducing the overall volume of optical systems. They can also provide ideal diffraction-limited focusing performance, while avoiding some problems associated with refractive lenses such as aberrations.

Read more | >

AI tool generates video from brain activity

“Alexa, play back that dream I had about Kirsten last week.” That’s a command that may not be too far off in the future, as researchers close in on technology that can tap into our minds and retrieve the imagery of our thoughts.

Researchers at the National University of Singapore and the Chinese University of Hong Kong reported last week that they have developed a process capable of generating video from . The research is published on the arXiv preprint server.

Using a process called imaging (fMRI), researchers Jiaxin Qing, Zijiao Chen and Juan Helen Zhou coupled data retrieved through imaging with the deep learning model Stable Diffusion to create smooth, high quality videos.

Read more | >

I’m shocked! Shocked, I tell you

Oh hey, AI enthusiasts and futurism fans! I’d love to share with you an article I recently wrote on my Substack. It takes you on a journey from the ancient Greek device known as the Antikythera mechanism, all the way to the generative AI explosion of 2023, tracing the history of computation and AI.

For more than a decade, I’ve been writing about technology, society, and the future, aiming to provide thoughtful analysis and critical thinking on the latest trends and their implications. I’ve been following these topics for over 15 years, and I am enthusiastic about initiating a meaningful conversation with you about the changing world and its intersection with technology.

Well, not that shocked.

Read more | >