Google Preparing to Partner with Taiwan's MediaTek on Next AI chip, Information Reports

A signage for Google is displayed near their office in Beijing, China, 04 February 2025. (EPA)
A signage for Google is displayed near their office in Beijing, China, 04 February 2025. (EPA)
TT
20

Google Preparing to Partner with Taiwan's MediaTek on Next AI chip, Information Reports

A signage for Google is displayed near their office in Beijing, China, 04 February 2025. (EPA)
A signage for Google is displayed near their office in Beijing, China, 04 February 2025. (EPA)

Alphabet's Google is preparing to partner with Taiwan's MediaTek on the next version of its AI chips, Tensor Processing Units, that will be made next year, the Information reported on Monday, citing people involved in the project.

However, Google has not cut ties with Broadcom, the chip designer it has worked with exclusively on the AI chips over the past several years, the report said, citing an employee at the San Jose-based company.

Like Nvidia,. Google also designs its own AI server chips, which it uses for internal research and development and also rents out to cloud customers, Reuters reported.

This approach gives Google a competitive edge in the AI race by reducing its reliance on Nvidia, even as rivals like Microsoft-backed OpenAI and Meta Platforms have seen a surge in demand for Nvidia chips.

Late last year, Google rolled out its sixth-generation TPU in a bid to give itself and its Cloud customers an alternative to Nvidia's chips, which are the most sought-after processors in the industry.

Google chose MediaTek partly because the Taiwanese firm has a strong relationship with TSMC and charges Google less per chip compared to Broadcom, the Information report added.

Google spent between $6 billion and $9 billion on TPUs last year, according to research firm Omdia, based on Broadcom's target for AI semiconductor revenue last year.



Rise in 'Harmful Content' Since Meta Policy Rollbacks, Survey Shows

The logo of Meta is seen at the entrance of the company's temporary stand ahead of the World Economic Forum (WEF) in Davos, Switzerland January 18, 2025. (Reuters)
The logo of Meta is seen at the entrance of the company's temporary stand ahead of the World Economic Forum (WEF) in Davos, Switzerland January 18, 2025. (Reuters)
TT
20

Rise in 'Harmful Content' Since Meta Policy Rollbacks, Survey Shows

The logo of Meta is seen at the entrance of the company's temporary stand ahead of the World Economic Forum (WEF) in Davos, Switzerland January 18, 2025. (Reuters)
The logo of Meta is seen at the entrance of the company's temporary stand ahead of the World Economic Forum (WEF) in Davos, Switzerland January 18, 2025. (Reuters)

Harmful content including hate speech has surged across Meta's platforms since the company ended third-party fact-checking in the United States and eased moderation policies, a survey showed Monday.

The survey of around 7,000 active users on Instagram, Facebook and Threads comes after the Palo Alto company ditched US fact-checkers in January and turned over the task of debunking falsehoods to ordinary users under a model known as "Community Notes," popularized by X.

The decision was widely seen as an attempt to appease President Donald Trump's new administration, whose conservative support base has long complained that fact-checking on tech platforms was a way to curtail free speech and censor right-wing content.

Meta also rolled back restrictions around topics such as gender and sexual identity. The tech giant's updated community guidelines said its platforms would permit users to accuse people of "mental illness" or "abnormality" based on their gender or sexual orientation.

"These policy shifts signified a dramatic reversal of content moderation standards the company had built over nearly a decade," said the survey published by digital and human rights groups including UltraViolet, GLAAD, and All Out.

"Among our survey population of approximately 7,000 active users, we found stark evidence of increased harmful content, decreased freedom of expression, and increased self-censorship".

One in six respondents in the survey reported being the victim of some form of gender-based or sexual violence on Meta platforms, while 66 percent said they had witnessed harmful content such as hateful or violent material.

Ninety-two percent of surveyed users said they were concerned about increasing harmful content and felt "less protected from being exposed to or targeted by" such material on Meta's platforms.

Seventy-seven percent of respondents described feeling "less safe" expressing themselves freely.

The company declined to comment on the survey.

In its most recent quarterly report, published in May, Meta insisted that the changes in January had left a minimal impact.

"Following the changes announced in January we've cut enforcement mistakes in the US in half, while during that same time period the low prevalence of violating content on the platform remained largely unchanged for most problem areas," the report said.

But the groups behind the survey insisted that the report did not reflect users' experiences of targeted hate and harassment.

"Social media is not just a place we 'go' anymore. It's a place we live, work, and play. That's why it's more crucial than ever to ensure that all people can safely access these spaces and freely express themselves without fear of retribution," Jenna Sherman, campaign director at UltraViolet, told AFP.

"But after helping to set a standard for content moderation online for nearly a decade, (chief executive) Mark Zuckerberg decided to move his company backwards, abandoning vulnerable users in the process.

"Facebook and Instagram already had an equity problem. Now, it's out of control," Sherman added.

The groups implored Meta to hire an independent third party to "formally analyze changes in harmful content facilitated by the policy changes" made in January, and for the tech giant to swiftly reinstate the content moderation standards that were in place earlier.

The International Fact-Checking Network has previously warned of devastating consequences if Meta broadens its policy shift related to fact-checkers beyond US borders to the company's programs covering more than 100 countries.

AFP currently works in 26 languages with Meta's fact-checking program, including in Asia, Latin America, and the European Union.