Mar 16, 2021

ByteDance to start producing AI chips

ByteDance
AI
chipset
semiconductor
Tilly Kenyon
2 min
TikTok owner ByteDance has begun hiring employees for a possible push into semiconductors
TikTok owner ByteDance has begun hiring employees for a possible push into semiconductors...

Beijing based ByteDance has confirmed that it is making a ‘foray’ into AI chips and has established a research team. The company has posted at least a dozen job openings related to semiconductors, including software engineers in cities such as Beijing and Shanghai. 

ByteDance told Chinese business magazine Caijing they have established a team to research the development of artificial intelligence chips. The plan is still in its early stages and the company is focused on arm-based server-side chips, according to the source.  

Earlier this month, during the Chinese government’s National People’s Congress, the Chinese Communist Party pledged to boost spending and drive research into cutting-edge chips and artificial intelligence, in an effort to contend with the US for global influence in its five-year plan. 

China and AI chips 

Many smart/Internet of Things (IoT) devices that we use daily are often powered by forms of AI such as facial recognition cameras and voice assistants. Some devices are powered by vast data centres in the cloud whereas others process on the device themselves through an AI chip. AI chips (also called AI hardware or AI accelerator) are specially designed accelerators for artificial neural network (ANN) based applications.

According to IC Insights, a market research firm, last year in China $143 billion worth of chips were sold, but only $22.7 billion of these were produced in the country itself. Just $8.3 billion worth were produced by Chinese-headquartered companies. 

ByteDance rival Baidu recently raised around $2 billion for its artificial intelligence chip unit Kunlun. Currently Kunlun chips are mostly used by Baidu on smart electric vehicles and cloud computing. A source stated that Baidu has considered making the Kunlun unit a standalone company, with the objective of commercialising its chip design.  

"It has become part of the strategic layout for internet giants to make chips by their own, and ByteDance should have entered the sector earlier," said Liu Dingding, a Beijing-based independent tech analyst.

Share article

Jun 17, 2021

Facebook Develops AI to Crackdown on Deepfakes

Facebook
MSU
AI
Deepfakes
3 min
Social media giant, Facebook, has developed artificial intelligence that can supposedly identify and reverse-engineer deepfake images

In light of the large tidal wave of increasingly believable deepfake images and videos that have been hitting the feeds of every major social media and news outlet in recent years, global organisations have started to consider the risk factor behind them. While the majority of deepfakes are created purely for amusement, their increasing sophistication is leading to a very simple question: What happens when a deepfake is produced not for amusement, but for malicious intent on a grander scale? 

 

Yesterday, Facebook revealed that it was also concerned by that very question and that it had decided to take a stand against deepfakes. In partnership with Michigan State University, the social media giant presented “a research method of detecting and attributing deepfakes that relies on reverse engineering from a single AI-generated image to the generative model used to produce it.” 

 

The promise is that Facebook’s method will facilitate deepfake detection and tracing in real-world settings, where the deepfake image itself is often the only information detectors have to work with. 

Why Reverse Engineering? 

Right now, researchers identify deepfakes through two primary methods: detection, which distinguishes between real and deepfake images, and image attribution, which identifies whether the image was generated using one of the AI’s training models. But generative photo techniques have advanced in scale and sophistication over the past few years, and the old strategies are no longer sufficient. 

 

First, there are only so many images presented in AI training. If the deepfake was generated by an unknown, alternative model, even artificial intelligence won’t be able to spot it—at least, until now. Reverse engineering, common practice in machine learning (ML), can uncover unique patterns left by the generating model, regardless of whether it was included in the AI’s training set. This helps discover coordinated deepfake attacks or other instances in which multiple deepfakes come from the same source. 

 

How It Works 

Before we could use deep learning to generate images, criminals and other ill-intentioned actors had a limited amount of options. Cameras only had so many tools at their disposal, and most researchers could easily identify certain makes and models. But deep learning has ushered in an age of endless options, and as a result, it’s grown increasingly difficult to identify deepfakes.

 

To counteract this, Facebook ran deepfakes through a fingerprint estimation network (FEN) to estimate some of their details. Fingerprints are essentially patterns left on an image due to manufacturing imperfections, and they help identify where the image came from. By evaluating the fingerprint magnitude, repetition frequency, and symmetrical frequency, Facebook then applied those constraints to predict the model’s hyperparameters. 

 

What are hyperparameters? If you imagine a generative model as a car, hyperparameters are similar to the engine components: certain properties that distinguish your fancy automobile from others on the market. ‘Our reverse engineering technique is somewhat like recognising [the engine] components of a car based on how it sounds’, Facebook explained, ‘even if this is a new car we’ve never heard of before’. 

 

What Did They Find? 

‘On standard benchmarks, we get state-of-the-art results’, said Facebook research lead Tal Hassner. Facebook added that the fingerprint estimation network (FEN) method can be used for not only model parsing, but detection and image attribution. While this research is the first of its kind, making it difficult to assess the results, the future looks promising. 


Facebook’s AI will introduce model parsing for real-world applications, increasing our understanding of deepfake detection. As cybersecurity attacks proliferate, and generative AI falls into the hands of those who would do us harm, this method could help the ‘good guys’ stay one step ahead. As Hassner explained: ‘This is a cat-and-mouse game, and it continues to be a cat-and-mouse game’.

Share article