AI Weekly: Meta’s flashy, AI-powered vision of the metaverse brushes over concerns

Join the top executives today at the Data Summit on March 9th. Sign up here.

This week, in the shadow of Russia’s invasion of Ukraine, Meta (formerly Facebook) unveiled its ambitious plan for AI – powered by Metavers. During the presentation, CEO Mark Zuckerberg highlighted some of the company’s efforts, including an AI system that allows users to build virtual worlds by describing them through text and internal projects to build a universal speech translator and voice assistant that responds to voice, eye movements, and body language.

Few, if any, of the artificial intelligence technologies that Zuckerberg has introduced are fundamentally new. Systems that can create interactive 3D models from text have Existed To Sometime, As well as systems that can translate speech directly from one language to another – one of the projects that Meta announced during its press conference. Meta’s push towards systems that can learn to translate language using just a few examples is a duplication of the work that startups do including Hugging Face. And Amazon has long been testing artificial intelligence that can read body language, along with acoustic and linguistic cues, to better anticipate what users want.

It can be argued that the headquarters is located to surpass what has been achieved technologically to date, of course, given the well-funded R&D division and the variety of user data. (Last fall, headquarters estimated That it would spend at least $ 10 billion on Facebook Reality Labs, the lab tasked with creating its augmented reality and virtual reality products.

In this future, AI eliminates the language barrier, and intelligent assistants address users’ whims – transferring them to tropical beaches by voice command. If only.

This is undoubtedly a rosier platform than Facebook, which lost users for the first time ever during the last fiscal quarter of Meta. Countless reports including that of Meta Found It Its platforms made it possible disinformation and misinformation To bloom, which in some cases leads to Riots and violence. According to internal documents shared by a Facebook employee who became the revealer of Frances Hogen and her legal team, more than 95% of the shared hate speech on Facebook remains on Facebook – despite Meta’s repeated public claims to the contrary.

The optimized version of Meta for Towers is less problematic. But it brushes aside problems that have already begun to emerge in the virtual world, including abuse and cyberbullying. B recently A BBC investigation in which he wrote an impersonation of a 13-year-old girl, the digital avatar was harassed, called and chased after him. Mateh was recently forced Apply “Personal Border” in its virtual reality platform for the fight against lewdness.

Meta claims – and continues to argue – that its moderation policy, backed by artificial intelligence, will ensure that its metaverse remains a safe place to work and play. But given that the company has failed to achieve this at a more basic level, it is difficult to understand the prospect of a hassle-free meta-run run by headquarters.

AI seems unlikely to be Meta’s salvation. On the contrary, it often exacerbates or obscures the scope of the company’s moderation issues. In 2019, engineers at Meta According to the report Found that a management algorithm on Instagram (owned by Meta) was 50% more likely to ban black users than white users. recently Reporting Found that at some point, the hate speech recognition systems Meta used on Facebook aggressively identified comments that defamed white people more than attacks on other demographic groups. And also a Incorrect translation By Meta’s artificial intelligence software led to the arrest of a Palestinian in 2017 by the Israel Police.

Even assuming that headquarters overcome countless technical – and political – issues that plague its AI systems, many groups, especially those who do not speak common languages, are afraid of losing their grip on their culture if it is controlled solely by a corporation. While Shemta may see artificial-powered meteors as the solution to its biggest platform’s troubles – which, in addition to hostility and misinformation, includes Decrease in revenue from ads – This is far from certain that the public will adopt it. They also should not; It will take more than glittering and ambitious technological demonstrations to dispel the fears that the headquarters will thrive in the ways it has long been.

To cover artificial intelligence, send news tips to Kyle Wigers – and be sure to sign up for AI’s weekly newsletter and bookmark our artificial intelligence channel, The Machine.

Thanks for reading,

Kyle Wigers

Writes a senior team in artificial intelligence

The VentureBeat mission Is to be a digital urban square for technical decision makers to gain knowledge about transformative organizational technology and transactions. Learn more


Related Articles