Toggle light / dark theme

The chief technology officer of a robotics startup told me earlier this year, “We thought we’d have to do a lot of work to build ‘ChatGPT for robotics.’ Instead, it turns out that, in a lot of cases, ChatGPT is ChatGPT for robotics.”

Until recently, AI models were specialized tools. Using AI in a particular area, like robotics, meant spending time and money creating AI models specifically and only for that area. For example, Google’s AlphaFold, an AI model for predicting protein folding, was trained using protein structure data and is only useful for working with protein structures.

So this founder thought that to benefit from generative AI, the robotics company would need to create its own specialized generative AI models for robotics. Instead, the team discovered that for many cases, they could use off-the-shelf ChatGPT for controlling their robots without the AI having ever been specifically trained for it.

Additionally, iProov reveals a 295 percent yearly increase in biometric attacks that use novel face swaps. Using this technique, attackers generate synthetic images with AI, either replacing the face of the victim with a video of him to bypass liveness checks or with the face of the attacker itself.

But beyond deep fakes and other more rudimentary presentation attacks where identities are spoofed using masks or other presentation techniques, another type of attack is gaining ground: digital injection.

Salesforce is leading a financing round in Hugging Face, one of the most highly valued startups helping businesses use artificial intelligence, at a valuation north of $4 billion, according to two people with knowledge of the situation. The roughly $200 million funding round more than doubles the share price and private valuation of the New York–based company, one of these people said.

Salesforce is paying a high price for a piece of Hugging Face, which runs a service that helps companies store and use AI software, similar to the way GitHub lets developers store software code. The new funding valued the startup at more than 100 times its annualized revenue, a measure of how much revenue the company would generate over the next 12 months at its current rate, one of the people said.

High-resolution 3D retina images have specific markers that can indicate the risk of Parkinson’s in a person. A new AI program can identify these markers and tell whether or not you have the disease.

Although Parkinson’s disease (PD) is incurable, a non-profit National Council of Aging report suggests that early detection and treatment could help patients live a long and productive life even with the disease.

However, in reality, even by the age of 50, less than 10 percent of patients are diagnosed. In fact, most PD patients found out about the condition in their 60s, and by then, it is too late for any treatment to work effectively.

The report further says 40 percent of workers will need to polish their skills due to the implementation of AI.

Artificial intelligence (AI) won’t replace employees anytime soon. But people who use AI will replace people who don’t, said tech giant IBM in its report, which talks about the implications of AI in businesses.

Companies are rapidly introducing AI into their workings to free up their employees’ time so they can focus on issues that require their personalized attention. The thing about AI is that it will do exactly what you train it to do. So, the hyperboles around the latest technology snatching away people’s jobs and taking over humanity can calm down.

In its quest to develop AI that can understand a range of different dialects, Meta has created an AI model, SeamlessM4T, that can translate and transcribe close to 100 languages across text and speech. Available in open source along with SeamlessAlign, a new translation dataset, Meta claims that SeamlessM4T represents a “significant breakthrough” in the field of AI-powered speech-to-speech and speech-to-text.

“Our single model provides on-demand translations that enable people who speak different languages to communicate more effectively,” Meta writes in a blog post shared with TechCrunch. “SeamlessM4T implicitly recognizes the source languages without the need for a separate language… More.