Zhitong
2023.08.02 02:23
portai
I'm PortAI, I can summarize articles.

Alphabet-C's Speech Assistant Department: Partial Layoffs, Introducing AI Technologies such as Large Language Models

According to an internal email from Alphabet-C, the company announced a restructuring of its voice assistant division to prioritize advancements in artificial intelligence.

According to Zhītōng Finance, as part of the latest major restructuring of Alphabet-C's voice personal assistant division, Alphabet-C hopes to "boost" its existing 7-year-old intelligent assistant using the latest advancements in generative artificial intelligence. Peeyush Ranjan, the engineering vice president responsible for Assistant business at Alphabet-C, said in an email to employees on Monday that the recent personnel adjustments will include a small number of layoffs. Ranjan stated that the company will seek to apply Large Language Model (LLM) technology to Alphabet-C's voice-driven software Assistant, which is similar to Apple's Siri or Amazon's Alexa.

Ranjan said, "As a team, we need to focus on providing users with high-quality, essential product experiences. We also see the tremendous potential of generative artificial intelligence in transforming people's lives and the enormous opportunity to explore a boosted assistant driven by LLM technology."

Ranjan added that part of the assistant team has already started working. The laid-off employees will have 60 days to find other jobs within Alphabet-C. As part of the restructuring, executives announced several adjustments to the "Speech" team responsible for voice commands within the company. Francoise Beaufays, who previously served as the head of the Speech department, will now work under Sissie Hsiao, who is in charge of Bard and Assistant.

In another email announcing the departmental changes, Beaufays wrote, "This is an exciting moment for artificial intelligence, as almost every product requires world-class AI-driven voice." She stated that some members of the Speech team will focus their work on Bard.

Assistant executives stated that these changes will enable the department to develop "quickly and focused." Alphabet-C spokesperson Jennifer Rodstrom stated that Alphabet-C is "excited to explore how LLM can help us strengthen our voice assistant and make it even better." She wrote, "Hundreds of millions of people use Alphabet-C Assistant every month, and we are committed to providing them with a high-quality experience."

The rapid development of generative artificial intelligence is driving Alphabet-C to embed this technology into as many products as possible. Generative artificial intelligence can respond to text-based queries with intelligent and creative answers, and it can also convert text into images.

For older assistant software, this means frequent improvements. Alphabet-C's mobile and home devices, including Pixel smartphones and Nest smart speakers and devices, all use Assistant. It is also applied to smartwatches, smart displays, televisions, and cars through the Android Auto platform. In March of this year, Hsiao announced an organizational reform, emphasizing priority consideration for Bard. An internal document shows that Ranjan, who previously served as Vice President of Business, has been appointed as the department's Engineering Supervisor, overseeing more than 1,700 full-time employees.

Since the launch of OpenAI's ChatGPT at the end of last year, Amazon has also emphasized the importance of generative artificial intelligence and has incorporated it into Alexa products.

For Alphabet-C, which has dominated internet search for most of the past 20 years, what is even more significant is that ChatGPT and Microsoft's search engine Bing, which utilizes OpenAI models, have provided people with an alternative way to search for answers. Undoubtedly, this has brought tremendous competitive pressure to Alphabet-C.

Since its public release in March of this year, Alphabet-C has been rolling out updates for Bard. Last month, the company announced that with its latest LLM technology, Palm 2, it will expand to more than 40 languages in additional countries and include features such as audio responses.