According to Bloomberg's Mark Gurman, Apple is set to spend as much as $1 billion a year in a determined effort to catch up with the rest of the AI industry. In his most recent Power On newsletter, Gurman explains that Apple's senior management were caught by surprise by the current AI trend, and the company is spending big to get back in the game.
Key among those efforts are a suite of generative AI tools including a reworked Siri assistant, which could arrive as soon as next year. Gurman also says we can expect AI updates to Apple Music, Pages, Messages, and other first-party apps as well. Additionally, Apple is reportedly looking to integrate AI features into development tools such as Xcode, which could streamline the app-building process.
Finally, Gurman says Apple is unsure whether to run the AI processing on-device, in the cloud, or some combination of both. Each strategy has its advantages and disadvantages, but with Apple making such a fuss about being a privacy-first company, there's certainly a lot at stake with how it rolls out AI.
Everyone's favorite time-wasting app Instagram is working on an AI-powered sticker-creation feature. As Engadget reports, Instagram head Adam Mosseri has shown off a yet-to-be-released feature that lets a user grab the subject of an image and then turn it into a virtual sticker that can be placed on top of other content.
In his demonstration, Mosseri noted that users will also be able to create stickers from other people's eligible images. Mosseri did not say when the feature will roll out, though it’s expected soon.
A team of researchers at IBM have developed NorthPole, an AI chip that can run image recognition tools more than 20 times faster than commercially available silicon.
The IBM Research unit, which is made up of engineers and computer scientists, say they have arranged the NorthPole's memory blocks and CPUs in such a way as to make communications more efficient than comparative chips. The researchers note the chip's architecture was inspired by how the human brain is able to pass information across different regions.
The IBM team says their chip can outpace silicon from Nvidia and other chipmakers, but they say NorthPole is not well-suited to training or running LLMs, thus limiting its long-term commercial viability.
The Wall Street Journal profiles several young entrepreneurs who have decided to forgo college and instead focus on building their own AI companies.
The piece talks to 19-year-old Govind Gnanakumar, founder of AI startup Automorphic, which was accepted into the prestigious accelerator program, Y Combinator. 20-year-old Kevin Lu, a fellow college dropout, has established Sweep AI, a company for automating programming tasks.
There are some dissenting voices when it comes to the opportunity that AI presents;. Tricia Martinez, an executive at accelerator Techstars, believes that using AI as a branding tool will not be effective in the long term. Despite such concerns, many entrepreneurs are not deterred. Jay Dang, the 21-year-old founder of FlowGPT, says that college is always a backup option if things don’t work out in the AI industry.