Apple Aiming to Cash in Generative AI Hype and Get Succeed

Apple (AAPL) – Get Free Report has long sought to distinguish itself from its competitors. The internet behemoth, whose leaders avoided using the word "artificial intelligence" during the company's developer conference in June, wants to build buzz rather than ride on existing enthusiasm.

That is not to say that Apple does not employ AI. It simply refers to it as machine learning, which is a subset of the wider AI category. The company's most recent product, the spatial computing Vision Pro, includes machine-learning technologies. Machine-learning algorithms are also present in everything from Face ID to battery consumption statistics, photographs, keyboard suggestions, and maps on Apple iPhones.

Nonetheless, Apple is hesitant to enter the generative AI game that its competitors Google, Microsoft, and Meta have been pushing so hard since the debut of ChatGPT last year. While rival digital behemoths have been engaging in an AI arms race to build and deploy consumer-facing generative AI, Apple has openly said that it is satisfied to employ AI technology without establishing a competitive Large Language Model (LLM).

But, behind closed doors, Apple has been working on the same generative AI technology.

According to Bloomberg, Apple has been developing an Ajax framework to handle LLMs; the business has also developed and internally deployed a service comparable to ChatGPT that some of its developers have dubbed "Apple GPT."

Next year, the business plans to launch a consumer-facing generative AI platform.

Apple did not reply to a request for comment from TheStreet.

What Apple Does Differently

Even though Apple is late to the race, the product that will be introduced to customers will most likely be far tighter than the ones that were pushed into deployment by concerned executives at Google and Microsoft.

Apple is the world's largest corporation. According to Warren Buffett, the explanation is that Apple has established an extraordinarily powerful and dedicated consumer base.

"Apple has a position with consumers where they're paying US$1,500 for a phone and US$35,000 for having a second car," Buffett stated in May. "And if they had to choose between giving up their second car and giving up their iPhone, they choose the second car." It's incredible."

Apple's choice not to enter the generative AI arms race was deliberate. That modest and sluggish approach to AI model development cannot potentially harm the eventual outcome.

Recent research discovered that ChatGPT's behaviour altered dramatically between March and June, and its performance dropped; GPT-4 was strong at recognizing prime numbers in March, but its accuracy dropped to roughly 2% in June.

"Absolutely stunning instability," tweeted AI specialist Dr Gary Marcus. "Who in their right mind would rely on a system that can be 97.6% correct in March and 2.4% correct in June?" Important outcomes. Take note if you want to rely on LLMs."

Meanwhile, Google's Bard was harshly criticized only a few months ago by Google's developers, who labelled the chatbot a "pathological liar" and "worse than useless."

And, despite becoming the fastest-growing site in history (until Meta released Threads, of course), chatbot usage has begun to decline. According to Similarweb, traffic fell 10% in June. Meanwhile, paid consumers have been complaining about a sudden drop in quality from the chatbot.

People now have access to AI models that are far from ideal. Instead of exposing people to a potentially harmful work-in-progress, Apple chose a more careful, comprehensive approach.

And, while only time will tell how wonderful their upcoming model will be, as long as it avoids the apparent shortcomings of the existing versions, Apple GPT may be able to sweep the competition.