Google reveals AI updates as it Vies with Microsoft

 Advanced Tech rolls on.  In China,  Baidu's Ernie makes a AI  play. 

Google reveals AI updates as it vies with Microsoft    in the BBC

Google has announced it is rolling out generative artificial intelligence (AI) to its core search engine.

The move comes after Microsoft incorporated GPT-4 into its Bing search engine earlier this year.Search Generative Experience - which will be part of Google - will craft responses to open-ended queries, the company said.

However, the system will only be available to a limited number of users and is still in "experimental" phase.

"We are reimagining all of our core products, including search," said Sundar Pichai, the boss of Google's parent company Alphabet. Additionally, the company announced a new feature on Google's Android system will proactively warn users about unknown AirTags, tiny devices developed to track personal items like keys and wallets.

The technology giant said the "unknown tracker alerts" would go live this summer.

The announcement came after Apple and Google said last week they were working together to address the problem.

Last year two women sued Apple over AirTag stalking.

Women who have been tracked using the devices told the BBC last year that not enough was being done to prevent misuse.

Apple AirTags - 'A perfect tool for stalking'

Google made the announcement at its annual developer conference, where leaders of the company touted their latest advancements in artificial intelligence and new hardware offerings, including a $1,799 (£1,425) phone that opens and closes like a book.

The company said it was removing the waitlist for "Bard", its experimental, conversational, chat service, which will be rolled out in English in 180 countries and territories.

It also said the chatbot would soon be able to respond to prompts with images as well as text.

Google has been under pressure to burnish its artificial intelligence offerings, after the runaway success of rival chatbot ChatGPT, which is funded by Microsoft.

A previous attempt to show off its credentials in the field, in February, ended in embarrassment, after it emerged that - in an advert intended to illustrate its capabilities - Bard had answered a question incorrectly.

The incident wiped $100bn (£82bn) off parent company Alphabet's share value - an indication of how keenly investors are watching how the tech giants' AI ventures play out.

Microsoft is deploying ChatGPT technology into its search engine Bing, after investing heavily in the company that developed it, OpenAI. Chinese tech giant Baidu also has a chatbot, called Ernie.

Chirag Dekate, analyst at Gartner, said Google remained an industry leader and was well poised to benefit in the interest in AI.

"Google has the tools to dominate the AI battles, the perennial question is - will they?" he said.

Early Looks at Ernie

 No direct testing seen yet.

Baidu Unveils Generative AI Ernie Bot With Pre-Recorded Videos, Disappointing Investors,   

 Eric Hal Schwartz in Voicebot.ai

Baidu introduced its generative Ernie Bot in a series of video presentations that promised a lot of features without actually demonstrating them live. The scripted nature of the event led to a brief dip of its Hong Kong shares by nearly 10%, though planning ahead did ensure Baidu avoided a gaffe like the incorrect answer to a question produced by the Google Bard generative AI chatbot when it was first revealed last month.

BAIDU ERNIE BOT

Baidu CEO Robin Li introduced Ernie, which stands for Enhanced Representation through Knowledge Integration, at the event. He played several videos showing Ernie performing math, generating images and videos from text prompts, and speaking in multiple Chinese dialects. The AI performed flawlessly in the video, but investors were apparently eager to see how it would perform in the wild, leading to the stock dip. Ernie is not widely available, with only those who received invitation codes from Baidu able to interact with the bot. Companies interested in employing Ernie for enterprise purposes must apply for the chance to link their software to Baidu’s cloud platform hosting the AI model.

Li admitted that Ernie is still not perfect, which might be why the company went for videos over live demonstrations. Li said Ernie had to be released “because the market demands it.” That seems to be backed up if Baidu’s boast of 650 companies in the Ernie ecosystem and 30,000 corporate user applications are to be believed.

Despite investor disappointment, Baidu has been working toward conversational generative AI for some time. The company has engaged in AI research for all kinds of products, including a virtual human named Xijiajia capable of painting a picture of a cat upon request. Xijiajia is Baidu’s digital human ambassador, one of several such virtual beings produced by Baidu, including the “interactive virtual idol” DuXiaoxiao. Some of the demand may also be a result of restrictions on American generative AI chatbots. Chinese social media platform WeChat banned ChatGPT last year, and strict rules on deepfakes have set up a potentially tricky regulatory system for any AI that can mimic specific people’s way of talking.

“Baidu envisions a future where we join forces with all to drive the evolution of AI, empowering every individual with access to state-of-the-art productivity tools and ensuring that the benefits of these advancements are shared by all,” Li said in a statement.  ... 

Baidu's Ernie Bot Debut

 Chinas Ernie Bot Doesn't do so well in first demo.  But the design of the western version is well known, so don't expect this lead to last. 

China’s ChatGPT rival Baidu Ernie is off to a rough start   From TechCrunch

Rita Liao@ritacyliao,  March 16, 2023

Baidu, China's search engine giant, introduced its Ernie Bot.

Following on the heels of GPT-4’s buzzy debut and the announcement of Microsoft 365’s AI makeover, Baidu, China’s search engine giant, introduced its Ernie Bot.

Since ChatGPT blew the world away, Baidu has been widely considered the closest Chinese candidate to build an equivalent to the OpenAI chatbot. Naturally, Ernie’s launch was much anticipated. On Thursday, Baidu CEO Robin Li gave a one-hour presentation on Ernie that only offered a small glimpse into the chatbot. The jury is still out on what Ernie can do and how it actually works.

For now, Ernie is only available for testing through invitation and others need to get on a waitlist. TechCrunch hasn’t tried it yet, so it’s unfair to draw any conclusions about Ernie’s capability.

But the public was clearly underwhelmed. Industry observers inside and outside China pointed to the fact that rather than showcasing Ernie through a live demo, Baidu opted for a lengthy presentation with pre-recordings of Ernie’s answers. The company’s shares slumped as much as 10% in Hong Kong following Li’s presentation.

In slides, Li presented the prompts for Ernie and its answers in five functional areas:

Literature writing: The user asked Ernie for advice on writing a sequel to the celebrated sci-fi novel “The Three-Body Problem.”

Business writing: The user asked Ernie to suggest names for a large-language model company that helps small-and-medium enterprises digitize.

Logic and reasoning: Ernie was asked to solve the famous “chickens and rabbits” math puzzle.

Chinese interpretation: Ernie was asked to define a traditional Chinese idiom and write a poem based on the phrase.

Multi-modal generation: Ernie was given a question (“Which city is most suited for implementing smart transport?”) and asked to perform “multi-modal” tasks based on the initial prompt, such as reading the answer out in a Chinese dialect and generating an image based on the text.

Baidu evidently did try to demonstrate what Ernie could achieve, and the answers were satisfactory; but still, investors weren’t impressed by the carefully orchestrated reveal. One can’t help but wonder if Baidu avoided a live demo due to its lack of confidence, and whether it had rushed the launch because of OpenAI’s impressive advancement. After all, even AI titan Google made a mistake in the demo of its conversational AI Bard.

Fangbo Tao, CEO and founder of AI startup Mindverse and a former AI scientist at Alibaba and Facebook, echoed the sentiment.    ....'

If Search is Really Reinvented

Seen from the retail direction, if Search is reinvented significantly, what will that mean?  Faster, more accurate?   Supply chains?   Search engine optimization certainly.    Mused about in pundits from Retailwire.  

Will ChatGPT reinvent search?  by Tom Ryan

Microsoft and Google both revealed plans last week to introduce their own ChatGPT-style product, joining the hype around the AI-driven chatbot technology.

Developed by San Francisco-based startup OpenAI and introduced last November, ChatGPT taps generative artificial intelligence to provide people with human-like responses to their questions. The technology also generates a wide range of digital text that can be repurposed in nearly any context, including in tweets, blog posts, term papers, poetry and even computer code.

Microsoft, which is investing $10 billion in OpenAI, will integrate ChatGPT technology with its Bing search engine and Edge web browser

Brad Smith. vice chair and president of Microsoft, said in a blog entry, “Today’s cutting-edge AI is a powerful tool for advancing critical thinking and stimulating creative expression. It makes it possible not only to search for information but to seek answers to questions. It can help people uncover insights amid complex data and processes. It speeds up our ability to express what we learn more quickly. Perhaps most important, it’s going to do all these things better and better in the coming months and years.

Google will soon offer an AI and natural language-enhanced version of its Google search engine called Bard, as well as a chatbot.

However, Microsoft created more buzz, having introduced a limited version of the new Bing available to some users this week to rave reviews and drawing predictions that Bing may threaten Google’s search dominance.

Microsoft’s CEO Satya Nadella said at a media event, “This technology is going to reshape pretty much every software category.”

In China, JD.com, Alibaba and Baidu confirmed they are working on ChatGPT-style projects, as well. Meta is also planning to release similar technology across various products.

Microsoft said its search engine includes technology that identifies and removes problematic content from the chat service. However, the A.I. technologies are known for often producing toxic content, including misinformation, hate speech and biases against women and people of color.  ... ' 

China's Baidu's ERNIE Plans to Compete

Have followed Ernie as an assistant style chatbot for some time, and saw chatbot style approaches like 'Ernie' in a visit to China.   Previously mentioned here.    Unclear yet how ERNIE compares to ChatGPT in abilities.  Here just a short overview.  

China’s Baidu reveals generative AI chatbot based on language model bigger than GPT-3  in TheRegister,  Tue 7 Feb 2023  

Probably bilingual bot will be called ERNIE outside China, Wenxin Yiyan within

Chinese web giant Baidu, which has “AI” in its name and has made AI the focus of its hyperscale cloud, has revealed it will launch a generative AI chatbot later this year.

Chinese media reported the launch and Baidu confirmed it to The Register.

“The company plans to complete internal testing in March before making the chatbot available to the public,” a Baidu spokesperson wrote.

The bot will be named “Wenxin Yiyan文心一言” or "ERNIE Bot" in English.

The spokesperson added that the bots are based on the Enhanced Representation through Knowledge Integration (Ernie) model first proposed in 2019. We were told Ernie “expands into a series of advanced big models that can perform a wide range of tasks, including language understanding, language generation (ERNIE 3.0 Titan), and text-to-image generation (ERNIE-ViLG).”

“What sets ERNIE apart from other language models is its ability to integrate extensive knowledge with massive data, resulting in exceptional understanding and generation capabilities,” the spokesperson added.

Just what the bot will be capable of is not known, but Baidu has over the years described its ambitions for something that sounds an awful lot like ChatGPT. As we noted yesterday, Baidu has discussed "proposed a controllable learning algorithm and a credible learning algorithm to ensure the model can formulate reasonable and coherent texts". It’s done the hard work to deliver such a system, having in 2021 detailed “ERNIE 3.0 Titan” – a pre-training language model with 260 billion parameters.

ChatGPT uses the 175-billion parameter GPT3 model.  ... ' 

Baidu Challenges GPT!

 AI thaws, and the Race is on.  Some details below, and at the link.   Had seen a demo of ERNIE and its use with Baidu before. 

Baidu to Launch Powerful ChatGPT Rival   By Ryan Daws | January 30, 2023 | TechForge Media

Categories: Applications, Artificial Intelligence, Chatbots, Companies, Machine Learning,

Ryan is a senior editor at TechForge Media with over a decade of experience covering the latest technology and interviewing leading industry figures. He can often be sighted at tech conferences with a strong coffee in one hand and a laptop in the other. If it's geeky, he’s probably into it. Find him on Twitter (@Gadget_Ry) or Mastodon (@gadgetry@techhub.social)

Chinese web giant Baidu is preparing to launch a powerful ChatGPT rival in March. .. 

Baidu is often called the “Google of China” because it offers similar services, including search, maps, email, ads, cloud storage, and more. Baidu, like Google, also invests heavily in AI and machine learning.

Earlier this month, AI News reported that Google was changing its AI review processes to speed up the release of new solutions. One of the first products to be released under Google’s new process is set to be a ChatGPT rival, due to be announced during the company’s I/O developer conference in May.

However, Baidu looks set to beat Google by a couple of months.

Bloomberg reports that Baidu will reveal its own AI-powered chatbot in March. The currently unnamed tool will be integrated into the company’s search product.  Powering the Baidu ChatGPT competitor is ‘ERNIE’ (Enhanced Language RepresentatioN with Informative Entities), a powerful AI model with 10 billion parameters.

Researchers have found that deep-learning models trained on text alone – like OpenAI’s GPT-3 or Google’s T5 – perform well for numerous problems, but can fall short on some natural language understanding (NLU) tasks when the knowledge is not present in the input text.

The first version of ERNIE was introduced and open-sourced in 2019 by researchers at Tsinghua University to demonstrate the NLU capabilities of a model that combines both text and knowledge graph data.

Later that year, Baidu released ERNIE 2.0 which became the first model to set a score higher than 90 on the GLUE benchmark for evaluating NLU systems. ... ' 

Reports on Computing, Big and Small

Training and More

New Records for the Biggest and Smallest AI Computers Nvidia H100 and Intel Sapphire Rapids Xeon debut on ML Perf training benchmarks   By SAMUEL K. MOORE  in Spectrum IEEE 

The machine-learning consortium MLCommons released the latest set of benchmark results last week, offering a glimpse at the capabilities of new chips and old as they tackled executing lightweight AI on the tiniest systems and training neural networks at both server and supercomputer scales. The benchmark tests saw the debut of new chips from Intel and Nvidia as well as speed boosts from software improvements and predictions that new software will play a role in speeding the new chips in the years after their debut.

Training Servers

Training AI has been a problem that’s driven billions of dollars in investment, and it seems to be paying off. “A few years ago we were talking about training these networks in days or weeks, now we’re talking about minutes,” says Dave Salvator, director of product marketing at Nvidia.

There are eight benchmarks in the MLPerf training suite, but here I’m showing results from just two—image classification and natural-language processing—because although they don’t give a complete picture, they’re illustrative of what’s happening. Not every company puts up benchmark results every time; in the past, systems from Baidu, Google, Graphcore, and Qualcomm have made marks, but none of these were on the most recent list. And there are companies whose goal is to train the very biggest neural networks, such as Cerebras and SambaNova, that have never participated.

Another note about the results I’m showing—they are incomplete. To keep the eye glazing to a minimum, I’ve listed only the fastest system of each configuration. There were already four categories in the main “closed” contest: cloud (self-evident), on premises (systems you could buy and install in-house right now), preview (systems you can buy soon but not now), and R&D (interesting but odd, so I excluded them). I then listed the fastest training result for each category for each configuration—the number of accelerators in a computer. If you want to see the complete list, it’s at the MLCommons website.  .... ' 

Baidu PaddlePaddle for China Industry

Baidu PaddlePaddle,   new to me, Here is a quick look.  Examining other potentials.

Baidu’s PaddlePaddle Spins AI up to Industrial Applications China’s ease with ubiquitous AI manifests at neural-net scales    By CRAIG S. SMITH   in Spectrum IEEE

TensorFlow, PyTorch, and Keras: Those three deep-learning frameworks have dominated AI for years even as newer entrants gain steam. But one framework you don’t hear much about in the West is China’s PaddlePaddle, the most popular Chinese framework in the world’s most populous country.

It is an easy-to-use, efficient, flexible, and scalable deep-learning platform, originally developed by Baidu, the Chinese AI giant, to apply deep learning to many of its own products. Today, it is being used by more than 4.77 million developers and 180,000 enterprises globally. While comparable numbers are hard to come by for other frameworks, suffice to say, that’s big.

Baidu recently announced new updates to PaddlePaddle, along with 10 large deep-learning models that span natural-language processing, vision, and computational biology. Among the models is a hundred-billion-parameter natural language processing (NLP) model called ERNIE 3.0 Zeus, a geography-and-language pretrained model called ERNIE-GeoL, and a pretrained model for compound representation learning called HELIX-GEM.

The company has also created three new industry-focused large models—one for the electric power industry, one for banking, and another one for aerospace—by fine-tuning the company’s ERNIE 3.0 Titan model with industry data and expert knowledge in unsupervised learning tasks.

Software frameworks are packages of associated support programs, compilers, code libraries, tool sets, and application programming interfaces (APIs) to enable development of a project or system. Deep-learning frameworks bring together everything needed to design, train, and validate deep neural networks through a high-level programming interface. Without these tools, implementing deep-learning algorithms would take a lot of time because otherwise reusable pieces of code would have to be written from scratch.

Baidu started to develop such tools as early as 2012 within months of Geoffrey Hinton’s deep-learning breakthrough at the ImageNet competition.

In 2013, a doctoral student at the University of California, Berkeley, created a framework called Caffe, that supported convolutional neural networks used in computer-vision research. Baidu built on Caffe to develop PaddlePaddle, which supported recurrent neural networks in addition to convolutional neural networks, giving it an advantage in the field of NLP.

The name PaddlePaddle is derived from PArallel Distributed Deep Learning, a reference to the framework’s ability to train models on multiple GPUs.

Google’s open-sourced TensorFlow in 2015 and Baidu open-sourced PaddlePaddle the next year. When Eric Schmidt introduced TensorFlow to China in 2017, it turns out China was ahead of him.

While TensorFlow and Meta’s PyTorch, open-sourced in 2017, remain popular in China, PaddlePaddle is more oriented toward industrial users.

“We dedicated a lot of effort to reducing the barriers to entry for individuals and companies,” said Ma Yanjun, general manager of the AI Technology Ecosystem at Baidu.

PyTorch and TensorFlow require greater deep-learning expertise on the part of users compared to PaddlePaddle, whose toolkits are designed for nonexperts in production environments.

“In China, many of the developers are trying to use AI in their work, but they do not have much AI background,” explained Ma. “So, to increase the use of AI in different industry sectors, we’ve provided PaddlePaddle with a lot of low-threshold toolkits that are easier to use so it can be used by a wider community.”

AI engineers normally don’t know much about industry sectors and industry-sector experts don’t know much about AI. But PaddlePaddle’s easy-to-understand code comes with a wealth of learning materials and tools to help users. It scales easily and has a comprehensive set of APIs to address various needs.  .... ' 

Global Driverless

No driver Taxi, but an in-taxi supervisor?

Baidu and Pony.ai become first robotaxi services to operate without safety drivers in Beijing

A supervisor must still be present in the vehicle, however,   in TheVerge By Emma Roth 

Baidu and Pony.ai have been given permission to operate their autonomous vehicles without safety drivers in Beijing, a first for robotaxi services in China (via CNBC). Although both companies now no longer need a staff member in the driver’s seat, they’ll still need a supervisor present somewhere in the vehicle.

Baidu and Pony.ai can’t operate throughout the entire city of Beijing just yet — they’re limited to a 60 square kilometer (23.1 square mile) area in Yizhuang, Beijing, the home of about 300,000 residents. While Baidu can deploy just 10 autonomous vehicles in the area, CNBC says Pony.ai can only operate four. Both companies have plans to expand the number of vehicles on the road (with Baidu shooting for 30), but it’s unclear how soon that will happen.  .... ' 

Baidu Makes Some AI Meta Humans

 Indeed. 

Baidu Unveils Its Metaverse of Custom Virtual Humans

ERIC HAL SCHWARTZ in Voicebot.ai

Chinese tech giant Baidu introduced a metaverse of virtual humans this week at Baidu Create 2021, in deliberate contrast to the presentation by Meta (formerly Facebook) CEO Mark Zuckerberg earlier this year. The digital world, named XiRang, is populated by humans and AI interacting through avatars built on the XiLing platform debuting in tandem with XiRang.

METAVERSE XIRANG

XiRang hosts a digital “Creator City” mimicking a physical city with original and recreated locales like the Shaolin Monastery. The virtual world is accessible through an app on computers and smartphones or with Pico virtual reality headsets. Baidu hosted CEO Robin Li and other executives at an AI conference within XiRang this week. Users with a Baidu service account can register on the app and build an avatar to travel around the virtual world, interacting with others by pointing their avatar at another and talking. XiRang, which means ‘land of hope,’ can support up to 100,000 users at a time right now, which Baidu emphasized because its metaverse is still under development.

According to Baidu CEO Robin Li, the metaverse will kick off a “golden era of AI over the next ten years” and the start of “man-computer symbiosis,” which will change the world. Li claims Baidu has more than a million employees working on AI and plans to quintuple that number in the next five years. “Creators are our companions in pursuing our dreams of technology, they are also one of the great driving forces of human progress,” Li said.  .... ' 

Quantum Computer Race in China

 Why and in what direction?

Competing Visions Underpin China’s Quantum Computer Race 

Alibaba builds their own qubits, Baidu remains quantum hardware-agnostic 

By  Craig S. Smith in Spectrum IEEE

China and the US are in a race to conquer quantum computing, which promises to unleash the potential of artificial intelligence and give the owner all-seeing, code-breaking powers.

But there is a race within China itself among companies trying to dominate the space, led by tech giants Alibaba and Baidu.

Like their competitors IBM, Google, Honeywell, and D-Wave, both Chinese companies profess to be developing "full stack" quantum businesses, offering access to quantum computing through the cloud coupled with their own suite of algorithms, software, and consulting services.

Alibaba is building solutions for specific kinds of hardware, as IBM, Google, and Honeywell are doing. (IBM's software stack will also support trapped ion hardware, but the company's focus is on supporting its superconducting quantum computers. Honeywell's software partner, Cambridge Quantum, is hardware agnostic, but the two companies' cooperation is focused on Honeyell's trapped ion computer.)

Baidu is different in that it is building a hardware-agnostic software stack that can plug into any quantum hardware, whether that hardware uses a superconducting substrate, nuclear magnetic resonance, or ion traps to control its qubits.

"Currently we don't do hardware directly, but develop the hardware interface," Runyao Duan, Baidu's head of quantum computing, told the 24th Annual Conference on Quantum Information Processing earlier this year. "This is a very flexible strategy and ensures that we will be open for all hardware providers."

Quantum computers calculate using the probability that an array of entangled quantum particles is in a particular state at any point in time. Maintaining and manipulating the fragile particles is itself a difficult problem that has yet to be solved at scale. Quantum computers today consist of fewer than 100 qubits, though hardware leader IBM has a goal of reaching 1,000 qubits by 2023.

But an equally thorny problem is how to use those qubits once they exist. "We can build a qubit. We can manipulate a qubit and we can read a qubit," said Mattia Fiorentini, head of machine learning and quantum algorithms at Cambridge Quantum in London. "The question is, how do you build software that can really benefit from all that information processing power?"

Scientists around the world are working on ways to program quantum computers that are useful and generalized and that engineers can use pretty much straight out of the box.

Of course, real large-scale quantum computing remains a relatively distant dream—currently quantum cloud services are primarily used for simulations of quantum computing using classical computers, although some are using small quantum systems—and so it's too early to say whether Baidu's strategy will pay off.

“We can build a qubit. We can read a qubit. But how do you build software that can really benefit from all that information processing power?"

In the past, Alibaba worked with the University of Science and Technology of China in Hefei, the capital of central China's Anhui province, which currently has the world's most advanced quantum computer, dubbed the Zuchongzhi 2.1, after China's famous fifth century astronomer who first calculated pi to six decimal places. The company is also building quantum computing hardware of its own.  ... ' 

Baidu Premium Smart Display

We examined and tested  the first Baidu smart assistants.

Baidu Aims at Premium Market With New X10 Smart Display

ERIC HAL SCHWARTZ on November 5, 2020 at 8:00 am

Baidu has launched its most expensive smart display ever in a bid for the more upscale smart home market. The Chinese tech giant is pitching the approximately $150 (RMB 999 Yuan) Xiaodu Smart Display X10 as the flagship device for its smart speaker and display catalog thanks to improved hardware and AI features operated through the Xiaodu voice assistant. Baidu showed off the smart display and its upgraded XiaoduPods earbuds at the 2020 Xiaodu Fall New Product Launch in Beijing.

PREMIUM DISPLAY

The X10 looks a bit like a tablet computer on top of a speaker and mounted on a stand. Unlike the plastic smart speakers and displays produced by Baidu before, the X10 has a metal skin and base, with a mesh cover for the speaker. The hardware is relatively powerful for its size, with 1280 by 800 resolution on the 10.1-inch screen, a five-megapixel HD camera, and 8W speakers The X10 includes voice and gesture commands for Xiaodu, which can recognize individual voices and carry on extended conversations without the user needing to repeat the wake word. Baidu worked with Silicon Valley-based design firm fuseproject to develop the look of the smart display.  smart devices functions that guarantee a seamless user experience for family members of all age groups. The comprehensive modifications include full-duplex continued conversation, gesture control, voiceprint recognition, and intelligent response capabilities, all empowered by Baidu’s leading AI technology.   .... ' 

Advances in Baidu Brain

Baidu is coming back, had looked at some of their earlier offerings and was not impressed, I get the impression they are now more seriously looking at the assistant market.  

Baidu Unveils Upgraded Baidu Brain AI and Smart Devices  By ERIC HAL SCHWARTZ 

Chinese tech giant Baidu showcased its latest creations at its yearly Baidu World conference. Much of the broad array of new products are powered by the new and improved Baidu Brain AI platform revealed at the same time, including robotic vehicles and the first smart television controlled with the Xiaodu voice assistant.  ... 

Baidu’s centerpiece of the event was its new robotic car, built with Baidu’s Apollo AI driving program. Apollo is embedded in the vehicle and is able to improve how well it performs not only in driving but in communications and interactions with passengers. Baidu has plans for automated taxis and other autonomous vehicles and is at the point of demonstrating them at a larger scale than most of its rivals. On the more domestic front, Baidu’s Xiaodu voice assistant has a few new homes, including its first smart television. The 86-inch TV is built with as much advanced tech as could be crammed within, including Xiaodu running the whole thing through voice and gesture commands. Xiaodu is also popping up in a new smart display and active noise-canceling headphones.  ... '

Baidu Announces Quantum Leaf

Quantum infrastructure a new term, but makes sense.

Baidu announces Quantum Leaf, a cloud-based quantum infrastructure servic   By Mike Wheatley  in SiliconAngle

China’s Baidu Inc. announced a new cloud-based quantum computing platform called Quantum Leaf today that it says is designed for programming, simulating and executing quantum workloads.

Baidu is one of a number of big technology firms racing to develop quantum computing, which relies on quantum-mechanical phenomena such as superposition and entanglement to perform computation. The technology promises to usher in a new era of computing and upend disciplines such as artificial intelligence, cryptography and physics. ... "

Baidu and Self Driving

Update on Baudu's involvement in self driving.

Baidu completes its computer for self-driving cars
It'll first be used for autonomous valet parking. ... '

Jon Fingas, @jonfingas in Engadget

Baidu Doing Drone Forestry Inspections with AI

Recall I worked with forestry management applications, so the need here rings true.  Note the system mentioned PaddlePaddle is open sourced.

Baidu’s deep-learning platform fuels the rise of industrial AI

PaddlePaddle lets developers build applications that can help solve problems in a wide range of industries, from waste management to health care.

by Baidu  ( In TechnologyReview.  This content was produced by Baidu.
It was not written by MIT Technology Review's editorial staff. ) 

AI is driving industrial transformation across a variety of sectors, and we’re just beginning to scratch the surface of AI capabilities. Some industrial innovations are barely noticed, such as forest inspection for fire hazards and prevention, but the benefits of AI when coupled with deep learning have a wide-ranging impact. In Southeast Asia, AI-powered forest drones have helped 155 forestry bureaus expand the range of forest inspections from 40% to 100% and perform up to 200% more efficiently than manual inspections. 

Behind these smart drones are well-trained deep-learning models based on Baidu’s PaddlePaddle, the first open-source deep-learning platform in China. Like mainstream AI frameworks such as Google’s TensorFlow and Facebook’s PyTorch, PaddlePaddle, which was open sourced in 2016, provides software developers of all skill levels with the tools, services, and resources they need to rapidly adopt and implement deep learning at scale. ... " 

Trick for Teaching AI the Meaning of Language

More advances in machine language learning and understanding.

Baidu Has Trick for Teaching AI the Meaning of Language
By MIT Technology Review

Baidu, Chinas closest equivalent to Google, recently outperformed Google and Microsoft in a competition known as the General Language Understanding Evaluation, or GLUE.

Chinese technology company Baidu outperformed Microsoft and Google in the General Language Understanding Evaluation, a benchmark for artificial intelligence (AI)'s comprehension of human language.

Researchers built the Enhanced Representation through kNowledge IntEgration (ERNIE) model specifically for the Chinese language, using a method that also makes the model better understand English.

The researchers built on the Bidirectional Encoder Representations from Transformers (BERT) model, which predicts and interprets the meaning of words by considering context before and after words simultaneously. BERT randomly hides 15% of words in a given passage of text and attempts to predict those words from the remaining ones, and Baidu modified this technique for ERNIE, hiding strings of characters rather than single characters; ERNIE also differentiates between meaningful and random strings to mask the right character combinations.

Baidu uses ERNIE to produce more applicable search results, eliminate duplicate stories in its news feed, and make AI-facilitated responses to requests more accurate.

From MIT Technology Review ... " 

Baidu and GLUE China Benchmarks

China and langage meaning.  A very key part of ultimately delivering useful conversation.

Baidu has a new trick for teaching AI the meaning of language in 7Wdata

Earlier this month, a Chinese tech giant quietly dethroned Microsoft and Google in an ongoing competition in AI. The company was Baidu, China’s closest equivalent to Google, and the competition was the General Language Understanding Evaluation, otherwise known as GLUE.

GLUE is a widely accepted benchmark for how well an AI system understands human language. It consists of nine different tests for things like picking out the names of people and organizations in a sentence and figuring out what a pronoun like “it” refers to when there are multiple potential antecedents. A language model that scores highly on GLUE, therefore, can handle diverse reading comprehension tasks. Out of a full score of 100, the average person scores around 87 points. Baidu is now the first team to surpass 90 with its model, ERNIE. .... "

Smart Speaker Stats

And much more stats at the link:

Amazon sold 10.4 million smart speakers in Q3 2019, three times more than Google’s 3.5 million devices  in Voicebot.AI

Alibaba lead Baidu in smart speaker sales in China with 3.9 to 3.7 million devices sold. Both companies sold more than Google during the quarter. Xiaomi was third in smart speaker sales in China for the second straight quarter.

Amazon Echo devices accounted for 36.6% of smart speakers sold worldwide up from 31.9% in Q3 2018 and up from 25.4% in Q2 2019.  .... " 

Baidu Surpasses Google in Voice

No surprise with volumes of people talking.   And further, since progress in this space can often depend on the volume of data that can be used to develop deeper insight into intelligent conversation,  this may be hard to stop.

Chinese search engine Baidu surpasses Google as 2nd-largest smart speaker vendor   By Patrick Hearn in DigitalTrends

In the western world, it’s easy to think of Google and Apple as the key players in the tech market. On a global scale, however, they have quite a few competitors — particularly the Chinese search engine giant Baidu, which made headlines this week when it became the world’s second-largest vendor of smart speakers.

According to the research firm Canalys, Baidu now owns 17.3 percent of the global market and made 4.5 million shipments in the second quarter of 2019. This represents a staggering year-on-year growth of 3,700%. In spite of this growth and success, Amazon still dominates the smart speaker market..... '