Schmidt’s removed speech deserves investors to read. What did it talk about?

Eric Schmidt

Introduction

Who is Eric Schmidt?

Eric Schmidt’s resume is as follows:

  • From March 2016 to November 2020, Schmidt served as chairman of the Department of Defense Innovation Advisory Committee and has been providing advice and suggestions to the Pentagon.
  • In 2015, after Google was reorganized into Alphabet Inc., Schmidt began to serve as Alphabet’s executive chairman and stepped down in 2017. But he is still a member of Alphabet’s board of directors until 2019 and is still a technical advisor to this day.
  • Schmidt served on Apple’s board of directors from 2006 to 2009.
  • 2001 to 2011: Google founders Riley Page and Sergey Brin, on the advice of venture capitalists John Doerr and Michael Moritz, recruited Schmidt to serve as Google’s CEO and concurrently president. One position. Schmidt stepped down as CEO in January 2011 and was replaced by Riley Page.
  • In 1997, Schmidt became president of Novell.
  • In 1983, Schmidt joined Sun Computer and later became Sun Computer’s chief technology officer (CTO) and corporate president.
  • Early years: Served as a researcher at Bell Labs, Zilog, and the Palo Alto Research Center, which was then under the jurisdiction of Quanlu.

Lecture at Stanford University

On August 14, 2024, Schmidt was invited to give a lecture at Stanford University. During this one-hour and 38-minute speech, he told the truth to the students in the audience with great sincerity, and told the students to keep it secret and not to spread it, and not to learn what he mentioned in his speech. Schmidt revealed many truths in today’s artificial intelligence circle during a nearly two hours long lecture and Q&A.

Originally, the organizer invited this elder in today’s technology industry to talk about artificial intelligence and other technological topics that everyone is most concerned about at present, but when he talked, he was completely unguarded and the whole article was in vernacular. Taboo provides an in-depth review of his inner words, as well as the development and trends of today’s global technology industry, the strength of various countries, the strength of major manufacturers, corporate operations and other sensitive topics.

The speech was immediately taken down

The video recording of this speech was originally posted on YouTube by the organizer afterwards. Schmidt’s speech attracted more than 40,000 views on YouTube in just two days and was immediately removed. However, because Schmidt talked too much and used unmodified words, it caused dissatisfaction among relevant people and caused controversy. In order to quell the controversy, the organizers had no choice but to remove the video of the speech from YouTube.

The removal of the wonderful speech does not mean there is a problem with the content, but because Schmidt was too honest, which caused dissatisfaction among stakeholders. The organizer had no choice but to give in.

The importance of this speech

Please go back and take a closer look at Schmidt’s experience. His identity is extremely sensitive and he is still an influential top heavyweight active in the American technology community. Therefore, his words have considerable credibility and weight. To a certain extent, it represents the current situation of the American technology community, and technology policy of U.S. government.

But everyone should know that Schmidt is not an “entrepreneur” for a long time, but a broker active in American politics. He advocates the theory of China’s technological threat, directly participates in the United States’ strategy of sanctioning and suppressing China’s technology industry, and makes profits from it. The most typical person.

Google’s decision to withdraw from China occurred during his tenure as Google’s CEO. This incident was very important in the technology community. Since then, Schmidt himself and Google have become the most famous anti-China leaders in the American technology community.

Internet links

It’s okay if it’s not removed the shelves, maybe no one will find out. As soon as it was removed, it immediately aroused the curiosity and interest of many people. I spent some time and found a video copy of this speech, as well as a verbatim transcript of the original speech. Please click on the two links below directly:

What did Schmidt say in this speech?

Taiwan

Two paragraphs of Schmidt’s speech specifically mentioned Taiwan. One is to comment on Taiwan’s software capabilities. He said, “Taiwan is a fantastic country whose software is terrible.” The second was when he discussed Taiwan’s superior manufacturing capabilities, mentioning that TSMC would allow Ph.D.s in physics to work in the factory in their first year, “Can you imagine American Ph.D. students working on the production line?” Schmidt’s comments were very clear. The strengths and pain points of technological development in Taiwan.

For an analysis of Taiwan’s software, please see my post of “Failed Taiwan Software Industry Policy” for details.

China

Schmidt is the most famous person in the American technology community and the earliest representative to advocate anti-China. He has been vigorously promoting the China threat theory on various occasions, including his role as an advisor to the Science and Technology Committee of the U.S. Department of Defense, where he serves.

Note: Google’s decision to completely withdraw from the Chinese market and close its offices in China was decided during Schmidt’s tenure.

But his argument is well-founded. In this speech, he named the medium and large countries in the world one by one, with the exception of China, which is qualified to become the only opponent of the United States.

Schmidt also said unabashedly: “China has risen rapidly, and the United States needs to increase investment to maintain its dominance. It is the U.S. government that bans the export of nVidia chips to China. This advantage has given the United States a few years’ lead.”

For China’s technology capability and ability, please see my post of “The hardware and software gap between China and US, is all China-made software and hardware possible?“

Geopolitics

In addition, Schmidt commented on the performance of major countries in the world in the field of AI. He believes that due to the EU’s regulatory restrictions, opportunities for technological innovation have been lost. Although France is working hard, Germany’s achievements are limited, and it is “powerless to recover.” Japan and South Korea are in our camp, but they are useless. In contrast, he believes that China is the main competitor of the United States, while India is an important country that has not yet been clearly analyzed.

Why are American afraid of TikTok?

“TikTok is not actually a social media platform, but more like a form of television. Each TikTok user in the United States uses the app for an average of 90 minutes a day and makes 200 videos, which is a very large amount of usage,” he said. .

“In the upcoming global elections, most misinformation will appear on social media, and social media companies’ current organizational capabilities are insufficient to effectively police this information. TikTok, for example, has been accused of favoring a certain kind of misinformation, although I There is no evidence.

He would not say that there is no evidence when he accuses Chinese technology companies on TV. And he didn’t stop there, but continued to share his true words.

“The government is trying to ban TikTok, and we’ll see if that actually happens.”

Note: TikTok, for this part please see my detailed analysis in another post: “TikTok, the rival of all social networks“

Promote plagiarism

he told Stanford students. “If TikTok is banned, I recommend that each of you give the following instructions to your language model (LLM): copy a TikTok, get all users and music, add my preferences to it, generate this program in 30 seconds, And publish it within an hour, if it doesn’t go viral, make some similar changes to the scenario in Python.

“What you would do if you′re a Silicon Valley entrepreneur, which hopefully all of you will be, is if it took off, then you’d hire a whole bunch of lawyers to go clean the mess up, right? But if nobody uses your product, it doesn′t matter that you stole all the content.”

AI is the modern arms race

“In order to make artificial intelligence more effective, I have become an arms dealer. Yes, it is legal,” he said.

Schmidt also served as chairman of the U.S. government’s Artificial Intelligence Commission, which led a series of crackdowns on China.

“I chaired an AI committee that looked at this issue very carefully. You can go read the report, it’s about 752 pages. I’ll just summarize it, we are currently leading, we need to stay ahead, and we need to A lot of money to do that.”

“The U.S. government basically banned all NVIDIA chips from China. Although they are not allowed to say what they are doing, this is indeed what they are doing to China.”

Employee work attitude

He said that Google currently lags behind OpenAI in artificial intelligence. He believes that employees are too idle, and pointed out that letting employees work from home or remotely and wanting to achieve a work-life balance is particularly the culprit. Not as desperate as TSMC and Musk’s companies.

Why did Google lose its AI dominance and was overtaken by OpenAI and Anthropic? He answered bluntly: “I am no longer an employee of Google, but it is obvious that Google has decided to win the competition by balancing work and life, going home early and working remotely. More important. And startups succeed because their people work hard. I’m sorry to be so direct, but that’s the truth.”

“I don’t like Musk’s personal behavior.” But he appreciated Musk’s work day and night. “We were chatting once, we were in Montana, and he was taking a flight that night to attend Midnight and X. ai conference.

He also affirmed the spirit of TSMC’s grassroots doctors: “TSMC has a rule that left a deep impression on me, that is, freshly graduated doctoral students, even outstanding physicists, have to work in the basement of the factory in the first year. Can you imagine an American Ph.D. in physics doing this?”

Regarding TSMC’s work culture, please see my post of “TSMC negative toxic corp culture and management style are detrimental to its future and growth” for details.

This comment immediately triggered an overwhelming response from the American technology community and Alphabet, forcing him to apologize immediately. It was also one of the main reasons why his speech was removed from the shelves.

Privacy is a myth

In the business world, he told the students directly, there is no data privacy at all. According to past experience, as long as you are successful enough, you can spend money to find the best lawyer to win the case for you. If you are unsuccessful, no one will sue you.

AI

AI is still in its infancy

We are still far from the popularization of artificial intelligence! The truth that no one wants to admit is: we are still far away from AGI (artificial general intelligence). The development of AGI also requires the simultaneous evolution of society and cannot be promoted by one company alone. Schmidt emphasized that he is looking forward to a new way of software programming that may more than double the productivity of the current way.

AI is a game for rich countries

Not only is this true for the entire business landscape, but at the national level, mastering artificial intelligence technology is also a game for rich countries, which requires a large amount of capital, technical talent and strong government support.

Schmidt also believes that the development of the AI ​​industry will make the rich get richer and the poor stay poor. This is true for individuals and the same for countries. Therefore, countries without technical resources must quickly join the AI ​​supply chain of powerful countries, otherwise they will miss out on this opportunity. In this important industrial revolution, the country’s competitiveness will be lost and it will become an insignificant economy.

AI ​​supply chain is huge

Money alone is not enough. “If you want to build a data center worth US$100 billion to US$300 billion, electricity will become a super scarce resource.”

In addition to money and energy, the industry itself is far from mature. Schmidt compared AI to electricity. Referring to past history, in the early days of the development of electricity, it did not instantly replace the productivity created by steam engines. It took 30 years until power plants and power supply networks were gradually completed, and the manufacturing industry using public electricity began to advance rapidly. Today’s AI is still in the “reaching for the low-hanging fruit” stage. The AI ​​industry still needs to establish a systematic industrial structure in order to achieve truly huge returns.

National level intervention is needed

To this end, he even went to the White House himself and said that the United States needs to build an alliance from now on: I told them that we need to be best friends with Canada, they have a lot of hydropower resources that our country does not have; another option is to let The Arab countries finance us, they have money.

This is not something that a single company can promote, but needs to wait for the synchronized evolution of the entire society. Schmidt emphasized that he is looking forward to a new way of software programming that may more than double the productivity of the current way.

Note: Regarding current artificial intelligence industry’s supply chain, please see my detailed analysis in another post: “Artificial intelligence benefits industries“

nVidia’s Monopoly

“Why is Nvidia worth $2 trillion while other companies are still struggling?” Schmidt asked the students present.

He pointed out that in fact, the entire AI industry chain, whether it is a startup or a giant, or even a top company like OpenAI, must rely on NVIDIA’s hardware resources. This dependence on the market makes the entire AI industry chain “unspeakably miserable for NVIDIA” , everyone has to act based on Nvidia’s face.

“If there were unlimited funds, anyone would choose the nVidia B200 because it is faster.” Schmidt said: “I talked with Su Zifeng of Super Micro for a long time, although they have already built the Rocm to replace the nVidia CUDA function, but the problem is that no one uses it.

This means that under Nvidia’s technology monopoly, even if competitors have the ability, it will be difficult to shake its market. He believes that Nvidia’s success lies in its ability to quickly adapt and meet the needs of AI development, which is in sharp contrast to the relatively slow progress of other companies.

He believes that companies that can invest heavily in Nvidia chips and data centers will be technologically ahead of unmatched competitors.

Schmidt also reflected on some of his misjudgments. For example, he once thought that NVIDIA’s CUDA was a stupid programming language, but now CUDA has become NVIDIA’s most powerful moat. Almost all large models must run on CUDA. Only NVIDIA’s GPU supports CUDA, which is a combination that other chips cannot shake.

However, do technical barriers like CUDA mean that other companies have no chance at all? Schmidt once talked to an entrepreneur and said that he was having a hard time switching: Should he choose Google’s TPU specially customized for AI, or use the more powerful Nvidia GPU to train his language model? If money was no object, he would definitely choose Nvidia’s B200 architecture, but the reality is that he must weigh between the two. In reality, Nvidia does bring a clear advantage.

Note: Regarding Rocm and CUDA, please see my detailed analysis in another post:

But Schmidt also showed his helplessness towards this monopoly: “Is there anyone here who does not have Intel chips in their computing equipment?” Unfortunately, no one raised their hands, and the foundation laid by Intel in the 1990s still dominates today. Following the market, it can be inferred that nVidia will still dominate the development of the entire AI trajectory for a long time to come.

He said that when talking to him, insiders of large companies claimed that they needed to invest 20 billion, 50 billion or even 100 billion US dollars in AI data centers based on nVidia chips. The construction cost of these data centers can be up to As much as $300 billion. He also said that he is a good friend of OpenAI executive Sam Altman.

Note: For the cost of artificial intelligence, please see my detailed analysis in another post: “How Much Does Generative AI Cost?“

Moreover, he told the students more clearly: Schmidt said, “If all these US$300 billion are invested in nVidia, you will know what actions to take in the stock market.”

Note: Regarding the company nVidia , please see my other post for detailed analysis:

Open source is unreasonable

The entire AI will also evolve towards a more polarized pattern. “Capital-intensive giants will spend huge sums of money to build data centers, and software manufacturers will also switch from open source to closed source to further consolidate the giants’ monopoly position.”

Regarding the data center, please see the analysis of my post of “Data center, a rapidly growing semiconductor field“

This almost breaks the beautiful illusion of universal AI and open source benefiting application development. This is the core value that many large open source models emphasize. “Open source is great, and my career and most of Google’s foundations are built on open source. But the big model is different. It’s too expensive, and no one can always be taken advantage of like Meta.”

Schmidt said frankly that open source means not making money, but open source means dedication. The open source model has made great contributions in the history of science and technology, but in the field of large models, the open source model is difficult to maintain. The cost of training is too high and many companies cannot afford it. Few can continue to invest like Meta. Even the upcoming third version of Mistral, which he invested in, is likely to be closed source because the training cost is too expensive and they need income.

Note: DeepSeek’s major breakthrough was announced through an article in his Washington Post column; Schmidt indirectly retracted his views in the article that China’s artificial intelligence lags far behind the United States and that he is not optimistic about open source code.

For DeepSeek’s breakthrough, please see my post of “DeepSeek routed the global AI and stock“

Eric Schmidt
credit: Wiki

Related articles

Disclaimer

  • 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.

Author: Andy Lin

Stock Investment Books Published English Blog English Simple Bio English Website Facebook Twitter Account LinkedIn Account 股票投資著作書籍 中文部落格 中文簡歷 中文網站 臉書 推特帳號 領英帳號

error: Content is protected !!