Tech
Sources Indicate ByteDance Intends to Build New AI Models on Huawei Chips.
(VOR News) – Three sources acquainted with the matter have revealed that ByteDance, the Chinese parent firm of TikTok, plans to build an artificial intelligence model that will be trained mostly with processors from Huawei Technologies in response to restrictions imposed by the United States.
This occurs while the massive social media platform is searching China for CPUs.
ByteDance has expanded its business to include domestic suppliers of chips used in artificial intelligence and has accelerated the development of its own processors since 2022, when the United States started limiting exports of potent AI chips, such as those made by market leader NVIDIA.
In the tech sector, artificial intelligence has become a vital component. Custom AI models—computer programs that utilise pattern recognition to make decisions—are being used by businesses in a variety of industries, such as gaming and e-commerce, to differentiate their goods and services.
The people who want to remain anonymous said that ByteDance’s next move in the competition for artificial intelligence will be to employ Huawei’s Ascend 910B chip to train a large-language AI model.
This is because the proposal involves ByteDance confidential information.
A fourth person claims that ByteDance also plans to create a new AI model, although they were unable to clarify if this model will incorporate Huawei processors or not.
ByteDance is already using the Ascend 910B to handle less computationally demanding inference tasks. Based on information from three different sources and pre-trained artificial intelligence models, these activities provide predictions.
Artificial intelligence model training is much more complex and requires massive amounts of data. Consequently, ultra-high-performance CPUs—like the high-end graphics processing units provided by Nvidia—must be used.
One of the people claims that the new model’s complexity and capabilities, which will be determined by its computational properties, will be inferior to those of ByteDance’s current artificial intelligence model, Doubao.
In response to a request for comment, ByteDance remained silent. Speaking on behalf of ByteDance, Michael Hughes, a TikTok representative in Washington, District of Columbia, stated that “the entire idea behind this is incorrect.” For the model, nothing novel is being created.
A LIMITED AMOUNT
One of the people claims that although ByteDance ordered over 100,000 Ascend 910B chips this year, the company had only gotten less than 30,000 of them as of July. This rate is not fast enough to suit the company’s needs.
Two of the individuals claim that ByteDance has been unable to set a timeframe for the new model because of a number of issues, such as low availability compared to Nvidia’s processors that are available in China and insufficient processing capacity.
ByteDance uses the most recent artificial intelligence technologies in its flagship large-language model, chatbot Doubao, which was redesigned and unveiled in August 2023. It is also used in a great deal of other applications, such as Jimeng, a text-to-video tool. This month, it unveiled two video-focused Doubao models in an effort to compete with OpenAI.
The amount of people using these kinds of apps has increased dramatically since the start of this year. ByteDance’s chatbot, which has over 10 million monthly active users, has become one of the most popular apps in China.
The three people claim that because of the growing focus on artificial intelligence (AI), ByteDance has grown to be one of Huawei’s biggest users of AI chips.
Furthermore, two of the people claim that it is the biggest buyer of NVIDIA’s H20 AI processor, which the US chipmaker customised for the Chinese market in reaction to trade restrictions. Two distinct sources claim that TikTok’s parent firm is also Microsoft’s most significant Asian customer for NVIDIA chips that are available via cloud computing.
A prior Reuters story claimed that ByteDance had budgeted $2 billion for NVIDIA processors the year before. NVIDIA declined to provide a statement. Microsoft declined to respond even after being contacted for one. Huawei declined to respond to Reuters’ requests for comments.
SOURCE: DN
SEE ALSO:
China’s Reticence on the U.S. TikTok Prohibition is Highly Indicative.
Apple and Meta Refuse to Sign the EU’s New AI Safety Agreement.
Tech
Tesla Is Unveiling Its Long-Awaited Robotaxi Amid Doubts About The Technology It Runs On
DETROIT — The long-awaited introduction of Tesla’s robotaxi at a Hollywood studio on Thursday night has sparked great expectations. Some analysts and investors believe the price is too high.
The business, which began selling software named “Full Self-Driving” nine years ago but still cannot drive itself, is set to demonstrate the so-called “Cybercab” car, which may lack a steering wheel and pedals.
The announcement comes as CEO Elon Musk attempts to persuade investors that his firm is more about artificial intelligence and robotics, despite struggling to sell its primary goods, an old range of electric vehicles.
Tesla Is Unveiling Its Long-Awaited Robotaxi Amid Doubts About The Technology It Runs On
Some observers anticipate that today will be a momentous day for the Austin, Texas-based startup, as it takes a significant step toward a long-awaited robotaxi service driven by AI.
Others who monitor self-driving cars believe Musk has yet to demonstrate that Tesla’s technology can travel safely without a human driver to intervene to avert accidents.
“I don’t know why the headlines continue to be ‘What will Tesla announce?’ rather than ‘Why does Tesla think we’re so stupid?'” said Bryant Walker Smith, a law professor at the University of South Carolina who specializes in autonomous vehicles.
He does not believe Tesla has the ability to demonstrate software and hardware that can operate without human supervision, even in a small area known as the driving system.
“We just haven’t seen any indication that that is what Tesla is working toward,” Walker Smith told me. “If they were, they would be showcasing this not on a closed lot, but in an actual city or on an actual freeway.”
Without a demonstrable advance in autonomous technology, Tesla will simply demonstrate a vehicle without pedals or a steering wheel, as many other firms have done, he said.
“The challenge is developing a combination of hardware and software plus the human and digital infrastructure to actually safely drive a vehicle even without a steering wheel on public roads in any conditions,” Walker Smith told Reuters. “Tesla has been giving us that demo every year, and it’s not reassuring us.”
Many industry analysts don’t expect much from the event either. While TD Cowen’s Jeff Osborne expects Musk to unveil the Cybercab and possibly the Model 2, a lower-cost electric vehicle, he does not see significant development in self-driving technology.
“We expect the event to be light on details and appeal to the true long-term believers in Tesla,” Osborne wrote in a note. Musk’s assertions about the availability of Full Self Driving, however, will be critical “given past delays and ongoing scrutiny” of the technology and Tesla’s less advanced Autopilot driver-assist software.
Tesla’s model selection is underperforming and is unlikely to be revamped until at least late next year, according to Osborne. Furthermore, he said that TD Cowen believes the “politicization of Elon” is hurting the Tesla brand among Democratic purchasers in the United States.
Musk has praised Republican presidential contender Donald Trump and supported other conservative initiatives. Last weekend, he joined Trump at a rally in Pennsylvania.
Musk has claimed for more than five years that a network of robotaxis is on its way, allowing Tesla owners to profit by having their cars transport passengers when they are not in use.
However, he has admitted that previous estimates for the adoption of autonomous driving were overly optimistic. In 2019, he projected a fleet of self-driving cars by the end of 2020.
However, Wedbush analyst Dan Ives, who is optimistic about Tesla stock, noted in an investor note that the company’s robotaxi event, called “We, Robot,” will mark a new chapter in Tesla’s growth.
Ives anticipates that Tesla will release more updates and specifics about the robotaxi, as well as improvements in Full Self Driving and artificial intelligence. He also wants a phased-in strategy for launching the robotaxis over the next year, as well as a Tesla ride-sharing app and demonstrations of technologies “designed to revolutionize urban transportation.”
Ives, whose team will attend the invitation-only event at Warner Bros. Studios, stated that he expected updates on Tesla’s Optimus humanoid robot, which the company intends to sell in 2026.
“We believe this is a pivotal time for Tesla as the company prepares to release its years of Robotaxi R&D shadowed behind the curtains, while Musk & Co. lay out the company’s vision for the future,” according to Ives.
Tesla Is Unveiling Its Long-Awaited Robotaxi Amid Doubts About The Technology It Runs On
The disclosure comes as US safety officials are looking into Full Self Driving and Autopilot, citing evidence that they have a weak system for ensuring human drivers pay attention.
Furthermore, the US National Highway Traffic Safety Administration ordered Tesla to recall Full Self-Driving in February because it permitted speeding and violated other traffic laws, particularly near intersections. Tesla planned to fix the concerns through an online software upgrade.
A motorcyclist was hit and killed by a Tesla employing Full Self-Driving in Snohomish County, Washington, near Seattle, in April, according to officials. The Tesla driver informed police that he was using the system and looking at his phone when the automobile rear-ended the motorcycle. Authorities reported that the motorcyclist was pronounced deceased on the site.
The National Highway Traffic Safety Administration said it is reviewing information from Tesla and law enforcement regarding the deadly crash.
SOURCE | AP
Tech
Amazon Says New Technology In Delivery Vans Will Help Sort Packages On The Fly And Save Time
Amazon has unveiled new technology that it claims would assist delivery drivers avoid sorting parcels at stops or manually checking to ensure they have the correct parcel for each delivery.
Amazon Says New Technology In Delivery Vans Will Help Sort Packages On The Fly And Save Time
Amazon’s retail president, Doug Herrington, said at a corporate event in Nashville on Wednesday that the technology, known as Vision-Assisted Package Retrieval, or VAPR, is meant to reduce the time and effort required by drivers to fetch parcels.
They plans to put 1,000 delivery vehicles with the technology on the road by early next year.
According to the firm, VAPR works as follows: When an Amazon delivery van arrives at a customer’s location, it projects a green “0” on parcels that need to be dropped off and a red “X” on those that should remain in the van.
According to the company, the system will also employ audible and visual clues to notify drivers when they have discovered the correct package.
Amazon Says New Technology In Delivery Vans Will Help Sort Packages On The Fly And Save Time
The Seattle-based corporation has been testing the new function with its Delivery Service Partners, which transport items across the country.
SOURCE | AP
Tech
OpenAI Plans To Establish Offices In Paris, Singapore, And Brussels To Facilitate Global Development.
(VOR News) – The news that OpenAI will be establishing new offices in a variety of locales all around the world was announced by the company in a post that was published on X.
New York City, Seattle, Paris, Brussels, and Singapore are some of the cities that fall into this category. As part of the company’s efforts to grow its operations across the globe, this announcement was made as part of those efforts.
In addition to the OpenAI offices that are already located in San Francisco, London, Dublin, and Tokyo, these new offices will also be constructed in a number of other places at a later date.
The ChatGPT maker, which is situated in San Francisco, has begun the process of extending its operations as a result of a number of factors, including a significant fundraising round with a total value of $6.6 billion.
There have been managerial changes and a reorganization plan for OpenAI.
According to the company, managing director Oliver Jay, who was formerly the chief revenue officer at Asana and the president of APAC and LATAM at Dropbox, would be responsible for supervising international operations and driving worldwide expansion from Singapore. Jay will be based in Singapore. Singapore is going to be Jay’s home base.
A representative of OpenAI shared with TechCrunch the information that the company has initiated the process of establishing a team in Singapore.
This information was provided to TechCrunch. The purpose of this team is to act as a central hub in order to provide help to clients and business partners that are located in the Asia Pacific region. It is currently in the process of hiring engineers, and it has plans to create an office in Singapore by the time this year comes to an end.
In addition, the company is currently in the process of recruiting engineers. In light of the fact that OpenAI’s Tokyo office was inaugurated in April, this will be the company’s second branch in the Asian region throughout the course of the current year.
According to a statement that was released by OpenAI, the number of users who are active on a weekly basis in Singapore has climbed by a factor of two since the beginning of the year.
This information was provided in OpenAI’s previous sentence.
According to the statement, the proportion of people in Singapore who use ChatGPT on a per capita basis is far higher than that of persons in any other country on the globe.
On top of that, OpenAI made the announcement that it would be forming a partnership with AI Singapore in order to broaden the availability of its artificial intelligence technologies across Southeast Asia.
It is through this relationship that OpenAI will be able to extend its sphere of influence. In 2017, the National Research Foundation (NRF) launched the Artificial Intelligence Singapore program, which is a national project dedicated to advancing artificial intelligence.
Supporting Singaporean businesses and research organizations in their efforts to implement artificial intelligence is the mission of this organization, which aims to provide assistance to these entities.
Sam Altman, chief executive officer of OpenAI, stated, “Singapore, with its extensive history of technological leadership, has established itself as a frontrunner in artificial intelligence, acknowledging its capacity to address significant societal challenges and enhance economic prosperity.”
Singapore has been a frontrunner in artificial intelligence. Singapore has ascended as a frontrunner in artificial intelligence.We are delighted to establish a cooperation with the government and the flourishing artificial intelligence ecosystem in the Asia-Pacific area following our operational expansion.
SOURCE: TCN
SEE ALSO:
US: A Judge Mandates that Google Allow Competing App Stores to Access Android
WhatsApp Now Features a “Mention” Tool for Status Updates and Stories.
Over The Planned “Link Tax” Bill, Google Threatens to Remove NZ News Links.
-
News1 week ago
The Biden Administration can go Ahead With Student Loan Forgiveness, Says a Federal Judge.
-
News1 week ago
Tesla Recalls 27,000 Cybertrucks Due To A Rearview Camera Issue
-
World1 week ago
Uber Hires Yandex Spinoff Ride-Hail and Autonomous Delivery With Avride
-
Tech1 week ago
Accenture and NVIDIA Collaborate to Enhance AI Implementation.
-
Tech1 week ago
Meta has started the Facebook Content Monetization Program.
-
Business1 week ago
McDonald’s Chicken Big Mac is Heading to the U.S. Next Week—for a Limited Time.