It is recommended that you read my previous article “OpenAI, the Generative Artificial Intelligence rising star and ChatGPT“, which mainly discusses generative artificial intelligence, so that you can have a preliminary foundation for the content of this article.
How big is the potential market?
Spending in the global generative AI market is expected to reach $42.6 billion by the end of 2023, growing at a CAGR of 32% to $98.1 billion by 2026, according to market analyst firm PitchBook Data.
There has been a monopoly
Rowan Curran, an analyst at Forrester Research, an IT research company, said that the generative artificial intelligence ecosystem, because the entry barriers are too high, is too large and diverse for any one supplier, even for ultra-large-scale suppliers. dominate the field. the
Curran said he expects barriers to entry in the generative AI market, such as the high cost of developing and training models, to loosen over time. This will enable more vendors to offer unique capabilities and fine-tune models that produce better results for the specific environments in which they operate, he said.
Microsoft (ticker: MSFT ) is trying to place ads in its Bing chat search engine, using words to generate images! Microsoft uses OpenAI technology to launch image creation capabilities on Bing. Microsoft’s mobile version of Bing adds voice search.
Because Bing search can obtain real-time updated network information (ChatGPT is not even the latest GPT-4 version). In order to avoid free riders from other manufacturers, competitors are unwilling to invest (obtaining daily real-time updates costs a lot, and few companies in the world have such capabilities), taking advantage of Microsoft. Microsoft has publicly threatened to cut off access to its internet search data, preventing it from licensing it to rival search engines, if rivals don’t stop using it as the basis for their own artificial intelligence chat products. In fact, what Microsoft refers to is Alphabet, because many media reports: GoogleAI researcher resigned after discovering that Bard used ChatGPT data.
Microsoft’s Azure cloud business is also a winner. We believe the Azure business will gain more market share this year because virtually all OpenAI and related traffic will happen on Azure.
Microsoft is moving quickly, based on the newly launched OpenAI GPT-4 module; it has released a software that can use artificial intelligence to collaborate with various software; the content product Microsoft 365 Copilot, including Office, can use pure text descriptions directly allow artificial intelligence to automatically generate relevant content for Office.
For example, after you write an article in Word, you can ask artificial intelligence to generate a PPT according to the content of the document, which can also add or use various special effects with text; and Excel, which is always a headache, can even directly subdivide numbers and assist in organizing related tables; Outlook can automatically summarize the content of emails; in the future, we can also look forward to launching programs related to project development and customer relationship management.
According to reports, Microsoft has launched a new artificial intelligence dialogue tool Security Copilot (aka Safe Copilot), which can help network security teams prevent hacker attacks and deal with them after encountering attacks. The Security copilot uses the GPT-4 language model developed by OpenAI, a professional artificial intelligence manufacturer, and professional data in the field of network security. Microsoft’s move is to use this tool to help staff quickly discover the connection between different pieces of evidence in a hacking attack, such as a suspicious email, a rogue software file, or the current status of an information system after an attack etc.
Alphabet (tickers: GOOGL and GOOG) is the biggest victim of this wave of ChatGPT, because ChatGPT may change users’ future Internet search behavior and directly threaten Alphabet’s largest source of revenue. You only need to look at the performance of Alphabet’s stock price since the beginning of 2023, and you can see that investors are worried about Alphabet’s prospects.
Alphabet’s Google is rumored to have invested nearly $400 million in artificial intelligence startup Anthropic, which is testing a competitor to OpenAI’s popular artificial intelligence chatbot ChatGPT.
Alphabet immediately pushed the chat robot Bard, directly pointing at ChatGPT. An internal memo from Alphabet’s CEO states that all employees are being asked to help test the Bard AI chatbot. The Alphabet version of ChatGPT is released, and does not support Chinese, and accuracy is still a problem.
Alphabet has opened Bard for trial use, but first vaccinated users, indicating that the answers provided by Bard may not be correct (everyone misunderstood, this is not because of poor Alphabet skills; all generative artificial intelligence software, including Microsoft and ChatGPT, have this problem , it’s just a matter of the media deliberately magnifying the problem by Alphabet).
But Alphabet encounters a new problem with generative artificial intelligence: because the cost of using generative artificial intelligence is 10 times that of traditional search, but even so, Alphabet cannot admit defeat. In fact, the LaMDA natural language model developed by Alphabet a few years ago is inherently designed for dialogue, while ChatGPT is a text-based design. From a purely technical point of view, ChatGPT is worse than LaMDA.
But Alphabet still has many moats, because it has a greater competitive advantage. Compared with ChatGPT, it has more data to be used as a basis, and has a wide range of computing infrastructure. Of course, Google Search also has a great market advantage on iOS and Android. The final formula of GPT-3 is 175 billion parameters and 250 billion characters (token) data, while Chinchilla, which belongs to DeepMind, uses 110 billion parameters and 500 billion tokens.
In order to join the artificial intelligence competition in Silicon Valley, the new large-scale language model LLaMA released by Meta (US stock code: META) is also a direct opponent of ChatGPT. The company stated that it will merge several teams to form a new product development team, led by Chris Cox, the current product director, to develop Facebook, Instagram, and WhatsApp integrated generative artificial intelligence products, which can automatically generate text pictures and other media.
Moreover, Meta is extremely efficient and has demonstrated the Make-A-Video multimodal algorithm to the outside world. Users only need to input scene descriptions to get short videos generated by artificial intelligence. The description does not have to be in words, pictures and videos can also be used as input data. According to Meta’s paper, this artificial intelligence model is trained using public databases.
In terms of the market value of Apple (ticker: AAPL) and its current leadership in the technology industry, Apple’s investment in artificial intelligence does not match the degree of respect it receives.
Apple’s investment in artificial intelligence and the report card it has handed over are mainly HomePod smart speakers that use the Siri voice assistant. But the HomePod smart speaker, as we all know, was Apple’s flop. Siri voice assistant, compared with Amazon’s Alexa voice assistant, or Alphabet’s Google Home voice assistant; the function is really much worse. It can only be said that it is barely better than the Cortana voice assistant that Microsoft has failed and has long since been abandoned by Microsoft.
However, Apple’s mergers and acquisitions in the field of artificial intelligence are very active. It has acquired many small artificial intelligence companies, and they have all been integrated into its products, software, or engines. In this part, Apple can be said to be more successful.
Amazon’s (ticker: AMZN) AWS and the startup Hugging Face, which is developing a ChatGPT rival product, have expanded their cooperation, and also said that Amazon has used ChatGPT in a variety of job functions, including writing code. But even so, Amazon is still among the top five technology companies, and its investment in artificial intelligence is relatively inferior. Except for the smart speakers using Alexa voice assistant, the achievements are really not too big.
On April 13, Amazon launched a platform service “Bedrock” on the company’s AWS, mainly for enterprise customers. Users can use generative artificial intelligence models such as AI21 Labs, Anthropic and Stability AI on the platform. In addition, Amazon also released its own large-scale language model “Titan”. This model will have two capabilities. One is a generative large-scale language model, which can be used to generate text, open-ended questions and answers, information classification, and key excerpts. It is an embedded large-scale language model that can convert text input into digital form and can help develop applications such as personalized search.
As the largest search engine in China, Baidu (ticker: BIDU) has launched its own Wenxin chatbot “ERNIE Bot (Wenxin Yiyan in Chinese)” in March 2023, and also provides an enterprise version, Baidu said that the number of enterprise users exceeded 70,000 in a short period of time. The day after the press conference, Baidu’s stock price rebounded sharply by 15% as soon as the market opened.
According to Baidu’s own estimate, it is about three months behind ChatGPT technically, but Baidu has an advantage that no one else has—that is, it has the world’s largest Chinese search database, it has a market share of more than 70% in China’s Internet search market, which is an invaluable treasure . This is also the biggest weakness of ChatGPT and the letter chat robot Bard, which are difficult to catch up with.
Note: According to the January 2023 market survey report conducted by statista; Baidu’s search engine has a global market share of less than 0.54%, Google’s is about 84.69%, and Microsoft’s Bing is 8.85%.
Moreover, ChatGPT was banned in China from the very beginning. As far as China is a huge market, Baidu is still ahead of other Chinese local technology competitors in the field of artificial intelligence; including Alibaba (ticker: BABA) and Tencent (ticker: TCEHY). So this is also the main reason why Wenxinyiyan may not be as good as ChatGPT technically, but Baidu’s share price has risen more than Microsoft in this wave of ChatGPT’s stock price rise.
As of the first quarter of 2023, Baidu’s stock price has risen by 27%, not only surpassing the 20% of Microsoft, the big financial backer behind ChatGPT, but also far better than the 7.5% of the US stock market. But in fact, Baidu’s fundamentals are not satisfactory, and its profit performance is mediocre. The latest financial report shows that its net profit in 2022 will be nearly 1.1 billion US dollars, a year-on-year decrease of more than 26%.
Adobe (ticker: ADBE) has long integrated Adobe Sensei’s artificial intelligence engine into its main products, using Adobe and its customer content and data to automate workflow, assist creative processes, and help businesses better understand their end customers.
Firefly is one of several AI-powered products and services that Adobe will launch in March 2023. Adobe uses its flagship software for areas such as corporate presentations, sales and marketing, and the company has launched its own generative artificial intelligence model.
The tool, called Firefly, aims to convert user prompts into image and text effects specifically for commercial use. Adobe said it plans to integrate the new tools into familiar workplace applications such as its best-known Photoshop and Adobe Illustrator.
One of the big selling points is that Firefly was trained using the company’s library of hundreds of millions of professional-grade, licensed stock images and expired copyright material, designed to prevent intellectual property infringement results, as well as other features tailored for business users. the
Some analysts see OpenAI’s Dall-E as a direct threat to Adobe’s core business. But it’s the features that Firefly offers that set the tool apart from general-purpose text-to-image platforms.
Through decades of experience in the enterprise market, Adobe has been able to understand how AI fits into business workflows; generating AI is a natural extension of that work.
Smaller software companies
The company C3.ai can be said to be the most popular target for US stocks and artificial intelligence investment in the first quarter of 2023. Exactly what is its charm, investors can refer to the introduction of the blog article I will publish in a few days: “C3.ai, a position successfully artificial intelligence company“.
Palantir (ticker: PLTR) is a big data and artificial intelligence vsoftware vendor mainly engaged in contracting government secret projects. Investors who want to know more about this company can refer to the description of my post: “What kind of company is Palantir?“.
SoundHound AI (ticker: SOUN) is a music recognition software, and Apple’s Shazam is a competitor. Currently Android, IOS, Symbian, and Windows Phone platforms are available. When users hear their favorite music but don’t know the title, they can use SoundHound to find it. Just point the microphone of the mobile phone at the sound source as much as possible or hum.
For most songs, SoundHound can obtain the title of the song from Midomi within 10 seconds, and display information such as the singer, album, album cover, as well as the singer’s tour date, Youtube video, and search for lyrics and purchases on the Internet. Some songs can also display lyrics information from LiveLyrics, and the lyrics will scroll synchronously with the song being played.
BigBear.ai (ticker: BBAI) is a software company that harnesses the power of artificial intelligence and machine learning to help governments and businesses make regulatory decisions that change markets and define outcomes.
Ashok Srivastava, chief data officer at Intuit (ticker: INTU), which owns TurboTax, the well-known tax preparation software, said his team is building a generative artificial intelligence language model for financial management, trained on years of interaction with business customers.
By combining company data with generative artificial intelligence and natural language technology, Intuit aims to make financial management tasks feel more natural by generating sound financial strategies based on user prompts, Srivastava said.
For example, Intuit recently released the beta version of its Email Content Generator solution, which is part of a suite of Mailchimp artificial intelligence (AI)-powered features. The launch aims at transforming email marketing for small and mid-size businesses. The solution is developed using Intuit Mailchimp’s existing generative AI tools and the company’s AI-driven expert platform. It is designed to help customers create email marketing campaigns based on industry, marketing intent and brand voice.
Investors who want to know more about Intuit, the company, can refer to the description of this post “Intuit, the leader in financial management software for individuals and SMEs”.
In the United States, no matter who you are, you will have a financial credit score, and FICO (ticker: FICO) is the company behind this credit score. It uses a complex set of algorithms, coupled with data from various cooperative manufacturers, to update consumers’ credit scores at any time. In short, its credit scoring system is a typical artificial intelligence algorithm, but when the company was founded and launched this system, the word artificial intelligence was not as popular as it is now.
If you want to know more about the investors of this company, you can refer to the description of my post “Fair Isaac, one of the most important listed companies in the US consumer finance industry“.
The Trade Desk
A key reason for The Trade Desk’s meteoric growth compared to the digital advertising market is its AI-powered demand-side platform (DSP), which allows advertisers to automatically buy and place ads based on their target audience.
The artificial intelligence engine used by The Trade Desk is called Koa, and Koa can play an important role in helping corporate clients provide relevant advertisements to their audiences. Koa analyzes data from around the web to ensure advertisers reach their audiences in the right medium in the most effective way. This helps explain why advertisers are now turning to a new generation of digital ad technology providers such as The Trade Desk, rather than placing ads through traditional channels.
The higher ROI from AI-driven advertising is why the programmatic display advertising market is expected to grow at an annual rate of approximately 36% over the next decade. GroupM, WPP’s media investment firm, predicts AI advertising will generate $1.3 trillion in revenue by 2032, up from an estimated $370 billion last year.
For those who want to know more about The Trade Desk, you can refer to the description of my post , “Deep Dive The Trade Desk“.
Upstart (ticker: UPST) is a financial start-up company that uses artificial intelligence technology to improve online loan efficiency and fraud, and is engaged in online loans. About this company, please refer to the description of my other article: “Artificial intelligence online lending platform Upstart“.
DraftKings (ticker: DKNG) is a US digital sports entertainment and gaming company. The characteristic of this company is the use of artificial intelligence technology to integrate it into the company’s lottery betting system.
On March 29, 2023, Synopsys (ticker: SNPS) launched new artificial intelligence tools for faster and better results at all stages of designing computing chips. It is designed to help engineers find errors in designs, test physical sample chips from partners, and increase the percentage of defect-free wafer completions once mass production begins.
In April 2023, Synopsys.ai was released, which is the industry’s first artificial intelligence-driven solution that comprehensively covers the most advanced digital and analog chips in the design, verification, testing and manufacturing processes. For the first time, engineers can use AI technology at every stage of chip design, from system architecture to design and manufacturing, and access the solution through the cloud.
Hardware infrastructure providers
Nvidia (ticker: NVDA) is so important that I separate it and plan to write another article. Please refer to “Revisiting Nvidia: The Absolute Leader in Artificial Intelligence, Data Center, and Graphics”. But readers can still refer to my two posts about this company a long time ago: “Data center, a rapidly growing semiconductor field” and “How does nVidia make money? nVidia is changing the gaming rules“; I believe it will be of great help to understand Nvidia.
Tesla (ticker: TSLA) is not just an electric car manufacturer, the company’s achievements in artificial intelligence are also worth mentioning to let investors know. Including its self-driving car, self-developed vehicle system, chips, humanoid robots, etc.; I won’t go into details here.
Intel (ticker: INTC ) , despite its poor performance of late, remains an integral provider of data centers and basic computing equipment for many businesses. If you want to know more about the investors of this company, you can refer to the explanation of this post “How does Intel make money? and the benefits to invest in it“.
Advanced Micro Devices (AMD)
The performance of AMD (ticker: AMD) in recent years is indeed impressive. It has indeed grabbed a large number of markets from Intel. Many companies are changing the high-priced Intel systems to better efficiency and prices. on the ultramicro system. Investors who want to know more about this company can refer to the explanation of this post “”How AMD makes money? A rare case of turning defeat into victory“.
Many friends must think that I am very disapproving of Broadcom (ticker: AVGO) as the main listed company of artificial intelligence. But people who hold this view are very wrong, because:
Broadcom is no longer just a semiconductor company. It already has a huge enterprise software department, and its turnover will account for more than 21% in the second quarter of 2023.
In addition to mainstream processors such as CPUs and GPUs that dominate computing power, artificial intelligence is also indispensable for hardware infrastructure and communication equipment, and Broadcom is the main supplier in this regard.
Customized special-purpose processors (ASICs) have also begun to be welcomed by many artificial intelligence customers, and Broadcom is the main ASIC supplier in the market.
If you want to know more about the investors of this company, you can read my previous post “How does Broadcom make money? Significant changes in business approach“
- “The artificial intelligence bubble in the capital market is forming“
- “Artificial intelligence investment trap“
- “Major artificial intelligence companies in US stocks market“
- “OpenAI, the Generative Artificial Intelligence rising star and ChatGPT“
- “Data center, a rapidly growing semiconductor field“
- “Consider Alphabet if you can buy one stock only to retire from it? “
- “How Microsoft makes money? What’s the next?“
- “How does nVidia make money? nVidia is changing the gaming rules“
- “Revisiting Nvidia: The Absolute Leader in Artificial Intelligence, Data Center, and Graphics“
- “How Adobe makes money? irreplaceable undocumented standard“
- “Artificial intelligence online lending platform Upstart“
- “Deep Dive The Trade Desk“
- “C3.ai, a position successfully artificial intelligence company“
- “What kind of company is Palantir?“
- “Fair Isaac, one of the most important listed companies in the US consumer finance industry“
- “Intuit, the leader in financial management software for individuals and SMEs”
- “How does Tesla make money?“
- “How AMD makes money? A rare case of turning defeat into victory“
- “How does Intel make money? and the benefits to invest in it“
- “How does Broadcom make money? Significant changes in business approach“
- The content of this site is the author’s personal opinions and is for reference only. I am not responsible for the correctness, opinions, and immediacy of the content and information of the article. Readers must make their own judgments.
- I shall not be liable for any damages or other legal liabilities for the direct or indirect losses caused by the readers’ direct or indirect reliance on and reference to the information on this site, or all the responsibilities arising therefrom, as a result of any investment behavior.