Experts warn of dire consequences from Meta’s GPT-4 rival.
Experts warn of dire consequences from Meta's GPT-4 rival.
Meta Develops AI System to Compete with GPT-4: Potential Problems and Concerns
Image Source: Brett Johnson / Unsplash
Facebook owner Meta is on a mission to surpass the capabilities of GPT-4, the impressive language model developed by OpenAI. Meta aims to launch its own artificial intelligence (AI) system in 2024, which it hopes will be “several times more powerful” than its recently released AI tool, Llama 2.
Meta’s plan is to create an AI model that can help other companies build services capable of producing sophisticated text, analysis, and more. With training set to begin in early 2024, Meta’s official release is expected to coincide with Google’s launch of its own large language model called Gemini.
To develop this advanced AI system, Meta is expanding its data centers and acquiring Nvidia H100 graphics cards. Mark Zuckerberg, Meta’s chief, also wants the technology to be open source, possibly to accelerate its adoption.
However, there are notable concerns surrounding the open-source approach. The accessibility of a powerful tool like this could have negative consequences if it falls into the wrong hands. The potential misuse could lead to an increase in viral disinformation and other damaging outcomes.
Moreover, questions arise regarding the sources of data used to train Meta’s AI model. Some competitors have faced criticism for allegedly plagiarizing artistic content and written materials to enhance their AI capabilities. Google, for example, has stated that everything on the internet is fair game for its Google Bard AI.
Meta’s track record on user privacy is also a cause for concern. Given Facebook’s numerous scandals, there is apprehension about entrusting personal data to Meta’s AI efforts. Previous worries arose when it was rumored that the company planned to launch AI chatbots with personalities, and this latest development is likely to intensify those concerns.
The current AI boom has raised so many potential negative consequences that prominent technologists have signed an open letter urging companies to hold off on developing AI systems more powerful than GPT-4 until society catches up with the implications. Unfortunately, these pleas seem to have fallen on deaf ears.
One cannot help but ask if any company is willing to prioritize societal cohesion and well-being over the vast profits that can be obtained from the AI industry. In Meta’s case, the answer seems to be a resounding no.
