Meta exec said company was behind on A.I., report says

Tech giants have been jockeying for an edge in AI for years, but perhaps one company is playing catch-up as the pressure intensifies. In a September note, a senior executive at Meta, Facebook’s parent company, mentioned that the company was still lagging behind in artificial intelligence.

“We have a huge gap in our tools, workflows, and processes when it comes to development for AI. We need to invest heavily here,” Santosh Janardan, Head of Infrastructure at Meta, said on an internal message board in September. Reuters reported on Tuesday.

The memo also said that the work of Meta for AI would require “a fundamental change in the design of our physical infrastructure, software systems, and approach to provide a stable platform.”

Nearly a decade after starting to invest in full AI research and streaming many of them Billions of dollars In this effort, Meta appears to be lagging behind in embracing hardware and software to support its AI ambitions, said the Reuters report, citing interviews with people close to the matter, company statements, and the memo.

The tech giant, which also owns Instagram and WhatsApp, is now trying to grow its AI prowess amid the rapid rise of OpenAI’s ChatGPT, a chatbot that can provide detailed answers to prompts in a matter of seconds. But in addition to shortcomings in its AI infrastructure, the Meta has been bogged down in the race by its focus elsewhere including on the metaverse, a virtual world that was one of CEO Mark Zuckerberg’s obsessions, according to Reuters.

In February, Meta announced a tool LLaMA that it will use for chatbots and other tools as it tries to compete with rivals Google and Microsoft. But with two rounds of layoffs and leadership changes, Meta is struggling.

Part of the problem identified in the Reuters report is Meta’s late adoption of graphics processing units, or graphics processing units, which can help speed up tasks, reduce the time needed to process massive amounts of data and make them suitable for some AI projects. Meta was also reportedly using its own custom reasoning chips, a process in which AI algorithms answer prompts and make judgments. Two people familiar with the matter reported that the approach to using dedicated GPUs and chips has slowed the company’s progress in the field of artificial intelligence. Meta eventually stopped using its own in-house chips and instead opted for chipmaker Nvidia’s GPUs, but by this time it was already far behind its rival, Google.

Also around this time, there was a change in its leadership for Meta on the AI ​​infrastructure side – including Janardhan taking over his role as Head of Infrastructure.

Meta is increasing its capital expenditures, or money invested in a company’s fixed assets such as equipment or land, through 2022, according to Reuters. That increase alone was $4 billion per quarter for most of last year and is expected to remain high through 2023 as well. This increase is due, in large part, to Meta’s attempt to expand its AI capability, Reuters reports.

But it’s the Menlo Park, California-based company that first started out on a large scale artificial intelligence research In 2013, she believes she can still break new ground when it comes to this technology.

“Meta has a proven track record of building and deploying advanced infrastructure at scale combined with deep expertise in AI research and engineering,” said a Meta spokesperson. luckEchoing what was stated in the Reuters report.

“We are confident in our ability to continue to expand the capabilities of our infrastructure to meet our needs in the short and long term as we introduce new AI-powered experiences to our suite of applications and consumer products.”