Apple teases new AI model that surpasses ChatGPT

Researchers from Apple (NASDAQ: AAPL) have developed a new artificial intelligence (AI) model designed to interpret context with minimal errors to foster natural conversation with users.

The new system, dubbed Reference Resolution As Language Modeling (ReaLM), allows the AI model to understand on-screen content that may appear ambiguous to other large language models (LLMs).

In their 15-page research paper, the researchers said Apple’s latest system demonstrated proficiency in understanding complex on-screen user commands, surpassing industry-leading offerings. Described as an “ideal choice” over tricky LLMs, the researchers describe ReaLM as the next frontier for conversational AI chatbots.

An example of the system’s capability can be seen in a test experiment conducted by the researchers. When asked to provide a list of pharmacies, users may ask the models to “call the bottom one” on the screen or the pharmacy on a particular street, posing a significant challenge for models incapable of understanding context.

Apple’s researchers relied on “parsed entities” to create a text-based representation of the screen, tagging the parts to create content. The first-of-its-kind solution is expected to be integrated with Siri and may run completely on the device, in line with Apple’s user privacy.

“We demonstrate large improvements over an existing system with similar functionality across different types of references, with our smallest model obtaining absolute gains of over 5% for on-screen references,” read the paper.

When placed side by side with ChatGPT-3.5 and other LLMs, Apple’s ReaLM outclassed older and larger models, rubbing shoulders with advanced models with its smallest version. The researchers surmised that ReaLM’s functionality is poised to provide “a true hands-free experience” for users of AI-powered voice assistants.

“Being able to understand context, including references like these, is essential for a conversational assistant that aims to allow a user to naturally communicate their requirements to an agent, or to have a conversation with it,” read the paper.

Apple’s delicate stance with AI

Despite taking the lead with AI and other emerging technologies with the release of Siri in 2021, the iPhone maker has fallen down the ranking in recent years. Experts attribute the decline to the cautious stance adopted by the technology company and its hard stance against cloud-based AI offerings.

Apple has since signaled an intention to join the foray, exploring high-profile deals with industry pacesetters Google (NASDAQ: GOOGL) and a string of media companies. In 2023, the company began experimenting with a generative AI model for employees to assist in product prototyping, recording impressive results along the way.

In order for artificial intelligence (AI) to work right within the law and thrive in the face of growing challenges, it needs to integrate an enterprise blockchain system that ensures data input quality and ownership—allowing it to keep data safe while also guaranteeing the immutability of data. Check out CoinGeek’s coverage on this emerging tech to learn more why Enterprise blockchain will be the backbone of AI.

Watch: How blockchain will keep AI honest