The Rise of ByteDance: How a Chinese Internet Giant is Revolutionizing Artificial Intelligence

The Power of Data

ByteDance, the Chinese internet giant behind popular apps like TikTok, Douyin, and Toutiao, has amassed a treasure trove of data from its hundreds of millions of users. With this wealth of information, the company has been improving its ability to recommend content and retain users. ByteDance’s data has also become the linchpin of its growing business in artificial intelligence (A.I.).

Investing in A.I.

ByteDance has invested heavily in the infrastructure needed to power A.I. systems, building vast data centers in China and Southeast Asia and purchasing advanced semiconductors. The company is also on an A.I. hiring spree, signaling its commitment to becoming a major player in the A.I. industry.

Concerns Over Data Use

ByteDance’s use of data has raised concerns globally, particularly in the United States, where lawmakers have tried to force the company to sell its U.S. operations due to concerns over national security and public opinion. However, in China, the data has helped ByteDance expand its business beyond social media and gain an edge in the global A.I. race.

A.I. and National Security

Chinese officials have pushed the country’s tech companies to pivot from entertainment apps to more strategic technologies like semiconductors, supercomputers, and A.I. ByteDance has embraced this mission, investing $11 billion in infrastructure like data centers and computer chips.

Workarounds for Chip Restrictions

The Biden administration has implemented rules to restrict Chinese companies from accessing certain chips, particularly those made by Nvidia. However, ByteDance has found ways to get the computing power it needs, likely by using data centers outside China and purchasing chips made by Chinese chipmakers like Huawei and Cambricon.

A.I. Applications

ByteDance’s A.I. efforts have led to the creation of popular apps like Doubao, a chatbot with 60 million users, and Volcano Engine, a business that lets other companies use ByteDance’s A.I. technologies. The company is also working with other companies, such as GAC Group and Mercedes-Benz, to integrate A.I. into their products.

Conclusion

ByteDance’s investment in A.I. has paid off, with the company becoming a major player in the industry. Its vast data and resources have enabled it to develop innovative A.I. technologies and applications. While concerns over data use remain, ByteDance’s focus on A.I. is likely to continue driving its growth and success in the years to come.

FAQs

Q: What is ByteDance’s business model?

A: ByteDance’s business model is centered around its popular apps like TikTok, Douyin, and Toutiao, which generate revenue through advertising and data analytics.

Q: What is ByteDance’s investment in A.I.?

A: ByteDance has invested heavily in A.I. infrastructure, hiring hundreds of A.I. researchers and engineers, and developing innovative A.I. technologies like chatbots and recommendation systems.

Q: What are the concerns over ByteDance’s data use?

A: Lawmakers in the United States have raised concerns over ByteDance’s data use, citing national security and public opinion concerns. However, in China, the data has been used to drive ByteDance’s growth and innovation.

Q: How has ByteDance adapted to chip restrictions?

A: ByteDance has found ways to get the computing power it needs, likely by using data centers outside China and purchasing chips made by Chinese chipmakers like Huawei and Cambricon.

A few years ago, Chef Robotics was facing potential death. “There were a lot of dark periods where I was thinking of giving up,” founder Rajat Bhageria tells TechCrunch of his six-year-old company. But friends and investors encouraged him, so he persevered.

The Grasping Problem

It all began when Bhageria did his master’s degree in robotics at UPenn’s famed GRASP Lab. He dreamed of the sci-fi promised world where robots did our housework, mowed our lawns, and cooked us five-star dinners. Such a world doesn’t exist yet because engineers have yet to fully solve the robotic grasping problem. Training the same robot to wash a wine glass without crushing it and a cast iron pan without dropping it is a difficult task.

When it comes to robotic chefs, “Nobody’s built a dataset of how do you pick up a blueberry and not squish it, or, how do you pick up cheese and not have it clump up?” he describes.

Saying No Leads to Yes

His original idea with Chef Robotics was similar to the long-list of the robotics startups that died: a robotic line for fast casual restaurants. That’s an enormous industry with a chronic employee shortage. “We actually had signed contracts. Like we had multimillion-dollar signed contracts. Obviously, we’re not doing this anymore. So what happened?” he said. “We essentially could not solve the technical problem.”

In those types of businesses, an employee completes an order by assembling all the varied ingredients necessary for each meal. These restaurants want robots to replicate that process because the alternative is to have dozens of robots dedicated to, and calibrated for, a single ingredient, some of which may only be used occasionally.

But Bhageria and team couldn’t build a successful pick-up-anything robot because the training data doesn’t exist. He asked his potential customers to let him install robots for one or two ingredients, gathering training data and building from there. They said no.

Then Bhageria had an epiphany. Instead of going bust trying to give existing customers what they wanted, maybe he needed different customers. “It honestly sucked, because I spent the last year and a half of my life trying to convince these people, these fast casual companies, to work up with us,” he recalled.

High Mix Manufacturing

It didn’t help that fundraising after 2021 was brutal. VCs were also looking at the graveyard. “We talked to dozens of different funds,” Bhageria said. “We just got rejected over and over.”

Bhageria was thinking of giving up. “You come home and are like, what am I doing in my life? Am I doing the wrong thing? Should I quit?” he remembered.

But he dug in and in March, 2023, raised an $11.2 million seed round led by Construct Capital, while also landing checks from Promus Ventures, Kleiner Perkins, and Gaingels.

Bhageria and team also found their perfect market, a part of the food industry known as “high mix manufacturing.” These are food makers that have many, many recipes, and make thousands of servings, but typically as meals or meal trays. For instance; salads and sandwiches or main courses and side dishes. These are meals used by airlines and hospitals, etc., or are frozen food meals for consumers.

Conclusion

Today, Chef Robotics has not only survived, it’s one of the few food tech robotic companies that is thriving. The startup, which recently raised a $23 million Series A, has 40 employees and marquee customers like Amy’s Kitchen and Chef Bombay. Dozens of robots installed across the U.S. have made 45 million meals to date, Bhageria says.

This compares to a graveyard of failed food tech robotics companies, including Chowbotics with its salad-making robot Sally; pizza delivery robot Zume; food kiosk robot Karakuri, and, more recently, agtech Small Robot Company.

FAQs

Q: Why did Chef Robotics almost die?
A: Chef Robotics almost died because it couldn’t solve the robotic grasping problem, which is a difficult task.

Q: What was the turning point for the company?
A: The turning point for the company was when Bhageria decided to focus on high mix manufacturing, a part of the food industry that has many recipes and makes thousands of servings.

Q: How did fundraising change for Chef Robotics?
A: Fundraising changed for Chef Robotics after it found its perfect market. It raised $11.2 million in a seed round led by Construct Capital, and later raised a $23 million Series A.

Q: What is high mix manufacturing?
A: High mix manufacturing is a part of the food industry that has many recipes and makes thousands of servings, typically as meals or meal trays.

Bluesky Updates App to Version 1.100: What’s New?

The latest update to Bluesky’s app brings a more comprehensive search page and new chat reactions to the table.

New Search Page: Explore

The search page is now known as “Explore,” where you can find trending topics, suggested accounts, and starter packs to jump-start stuff for you to follow on the app. This new Explore page still lets you search from the top, but now, the first thing you see under the search bar are a list of top trends and that may have a tag emphasizing how “hot” or “new” something is.

There’s also a shortcut at the top to edit interests that inform what you see on the Explore page, but you can hide that and adjust your interests later in Settings > Content and media.

New Chat Reactions

You can now hold down on a direct message and select an emoji to react to it with. The feature looks and works similarly to many other chat apps like iMessage.

Previously, holding down on a message brought up a menu on the bottom of the screen to translate, copy, or delete the message. Now, those options appear under a message when you press and hold on it.

What’s Next?

With these updates, Bluesky aims to make it easier for users to discover new content and engage with the app in a more interactive way.

Conclusion

The update to version 1.100 brings significant improvements to the search functionality and chat interactions, making Bluesky a more engaging and user-friendly platform.

FAQs

Q: What’s the new search page called?

A: The new search page is called “Explore.”

Q: How do I edit my interests on the Explore page?

A: You can edit your interests on the Explore page by clicking the shortcut at the top, or later in Settings > Content and media.

Q: How do I react to a message?

A: You can react to a message by holding down on it and selecting an emoji.

Q: What happened to the menu when I hold down on a message?

A: The menu is now located under the message when you press and hold on it.

Google’s Plan to Combine AI Models for a Universal Digital Assistant

DeepMind CEO’s Recent Podcast Appearance

In a recent appearance on Possible, a podcast co-hosted by LinkedIn co-founder Reid Hoffman, Google DeepMind CEO Demis Hassabis said Google plans to eventually combine its Gemini AI models with its Veo video-generating models to improve the former’s understanding of the physical world.

A Vision for a Universal Digital Assistant

“We’ve always built Gemini, our foundation model, to be multimodal from the beginning,” Hassabis said, “and the reason we did that [is because] we have a vision for this idea of a universal digital assistant, an assistant that … actually helps you in the real world.”

The Rise of Omni Models

AI Industry Trends

The AI industry is moving gradually toward “omni” models, if you will — models that can understand and synthesize many forms of media. Google’s newest Gemini models can generate audio as well as images and text, while OpenAI’s default model in ChatGPT can natively create images — including, of course, Studio Ghibli-style art. Amazon has also announced plans to launch an “any-to-any” model later this year.

Training Data for Omni Models

These omni models require a lot of training data — images, videos, audio, text, and so on. Hassabis implied that the video data for Veo is coming mostly from YouTube, a platform that Google owns.

YouTube as a Source of Training Data

“Basically, by watching YouTube videos — a lot of YouTube videos — [Veo 2] can figure out, you know, the physics of the world,” Hassabis said.

Conclusion

Google’s plan to combine its Gemini AI models with its Veo video-generating models is a significant step toward building a universal digital assistant. By leveraging training data from YouTube and other sources, Google aims to improve its models’ understanding of the physical world. As the AI industry continues to evolve, it will be interesting to see how these omni models shape the future of AI.

FAQs

Q: What are Gemini AI models?

A: Gemini AI models are multimodal models that can generate audio, images, and text.

Q: What is Veo?

A: Veo is a video-generating model developed by Google.

Q: Where does the training data for Veo come from?

A: The video data for Veo comes mostly from YouTube, a platform that Google owns.

Q: What are omni models?

A: Omni models are AI models that can understand and synthesize many forms of media, including images, videos, audio, and text.