Podcast source: Empire

Compiled & translated by: Deep Tide TechFlow

Guest: Wang Qiao

Host: Jason Yano

Original title: Claude Opus 4.5's Breakout Moment & Investing in 2026 with Qiao Wang

Broadcast date: January 12, 2026

Key points summary

This week, Qiao Wang was a guest on the program, discussing how artificial intelligence will redefine the meaning of being an investor in 2026. We delved into topics such as the breakthrough moment of Claude Opus 4.5, why Qiao chose to invest in Google, how to build an investment portfolio for 2026, and how to allocate time effectively in the age of artificial intelligence.

Summary of key viewpoints

The market sentiment is currently overly optimistic.
"About 40% of my portfolio is cash."
Crypto tokens are not very attractive, but there will always be opportunities in the market to focus on analyzing assets one by one, rather than simply looking at the overall asset class.
The stock and Bitcoin holding ratio is approximately 50/50.
My biggest investment is in Google, and I also hold shares in Tencent, which is a very good company.
One area currently being overlooked by the market is AI-driven biotechnology, which has enormous potential in the field.
By 2026, we may see some unicorn startups with only one or two people.
Some of the most successful AI startups are not companies like ChatGPT or OpenAI, but rather smaller businesses that utilize artificial intelligence.
The fundamental nature of a company's competitive advantage has not changed, but the competitive advantage in the software field is indeed weakening rapidly.
The code itself is no longer the bottleneck; the key is to design appropriate "prompts".
Gemini is undervalued by at least two orders of magnitude. "I would be willing to pay Gemini $2,000 per month."
I think Adobe could be the Google of this year because its current valuation is very cheap, and the market's view of it is somewhat similar to the view of Google back then.
Everyone should learn to code, and must learn to do so, as everyone can use automation tools to optimize certain aspects of their lives and work.
AI tools can make already efficient and capable people even more efficient and intelligent.
In terms of health, the most important factors are still three basic ones: diet, sleep, and exercise.

Qiao's 2026 Investment Portfolio

Yano: About a month ago you mentioned the current state of the market and expressed some concern. You also mentioned the dot-com bubble period, such as in 1996, when some people predicted that the market was about to peak, but the market continued to rise for three years.

Additionally, I saw one of your tweets discussing the relationship between forward P/E ratios and returns over the next ten years. You offered some counterarguments, but concluded by saying, "I'm still scared." I'm curious, what's your current view on the market?

Qiao:

My fear stems from an overly optimistic market sentiment. While I can't say everyone is, you can see many people on social media boasting about their investment gains, especially given the market volatility of the past week—this sentiment is undeniable. Quantitatively, market valuations are near all-time highs. Of course, some might argue that companies are now more profitable, have stronger competitive advantages, and have a more robust economic structure, coupled with the market-boosting effect of loose monetary policy, making current valuations reasonable. However, we cannot deny that market valuations are indeed at historical highs. Therefore, I haven't fully invested in risky assets. In fact, about 40% of my portfolio is in cash. While I might miss some opportunities if the market continues to rise, at least I can sleep more soundly.

Yano: I listened to interviews with Drucker and Mueller, and he mentioned, " In my 50-year investment career, this is the most difficult moment I have ever encountered in predicting market movements." If even Drucker cannot predict it, how can we ordinary investors possibly do so?

Qiao:

This is why I focus more on individual stock selection . If you look at the overall market, valuations are indeed high, but if you research individual companies one by one, you sometimes find very attractive investment opportunities. This way, even if the overall market declines, your fear will be reduced. During the dot-com bubble of 2000, while tech stocks underperformed over the next decade, small-cap and value stocks performed very well, with an average annual growth rate of 10%. I believe this is the best time for individual stock pickers.

Yano: What are your thoughts on investing in stocks and cryptocurrency tokens right now?

Qiao:

Currently, I don't think crypto tokens are very attractive , but there are always opportunities in the market, like in 2022. Many worthwhile investment opportunities emerged then. Therefore, I'm currently focusing on analyzing individual assets rather than simply looking at the overall asset class . From a holistic perspective, there are few particularly attractive assets right now, except for the US dollar. However, if you analyze assets individually, you might find some opportunities. As for cryptocurrencies, I still hold some Bitcoin and a small amount of tokens, but it's a small percentage of my overall investment.

Yano: If we try to break down your portfolio, 40% is cash, what percentage of the remaining 60% is stocks and cryptocurrencies? What is the percentage of Bitcoin and other crypto assets?

Qiao:

It's roughly 50/50 between stocks and Bitcoin. I still hold a significant amount of Bitcoin , but it's not my ideal portfolio because selling it incurs high taxes, and the proportion of cryptocurrency tokens is very small, less than 1%.

Yano: So what stocks do you currently hold?

Qiao:

My largest investment is Google, and I also hold shares in Tencent. Tencent is an excellent company; although its business is relatively low-profile, its fundamentals are very solid.

Yano: Do you think Tencent is better than Alibaba?

Qiao:

Currently, yes. Alibaba has a greater advantage in AI, but its retail business is highly competitive. I also hold some Amazon stock. I felt that market sentiment last year was very unfavorable for Amazon because all the tech giants' stock prices rose, while Amazon's performance was weak. Furthermore, I am very optimistic about Amazon 's investments in robotics. These investments are all long-term, with a ten-year target.

Yano: Amazon is indeed leading in robotics, and they may become the first major company to have more robotic employees than human employees.

Qiao:

Over the past five years, Amazon's human workforce has remained stable, but its robotic workforce has grown by 20% to 30%. They have significantly improved profit margins in recent years, despite modest revenue growth.

Yano: What are your thoughts on the medical field, such as Eli Lilly or other companies?

Qiao:

I hold some Eli Lilly stock, but it's a short-term trade, not a long-term investment, because my understanding of the pharmaceutical industry is limited. I don't know what will happen in ten years; I do know I might still be using Amazon in ten years, but there are just too many competitors in the pharmaceutical industry. While Eli Lilly has strong patent barriers, there are many cheap and effective grey market Chinese peptides that directly compete with Eli Lilly.

I think one area currently being overlooked by the market is AI-driven biotechnology . While everyone is focusing on robots, drones, and AI chatbots, I believe AI has enormous potential in the field of biotechnology.

Yano: I'd like to hear your thoughts on the automotive and fintech sectors. Do you own shares in Tesla or Rivian?

Qiao:

No, Tesla's valuation is too high. I don't know much about Rivian.

Yano: What about in the fintech sector? Like Robinhood and Coinbase?

Qiao:

From a valuation perspective, I find Coinbase more attractive, but neither company is a particularly outstanding investment. I hold Coinbase stock, while Robinhood's valuation is too high.

Yano: Do you think anyone in the West might develop a super app similar to "Mr. Know-It-All"?

Qiao:

Are you referring to a super app like WeChat that integrates messaging, payment, and social networking functions? I think there's already a similar trend in the financial sector, like Robinhood. But I don't see the possibility of a super app like WeChat that comprehensively covers all aspects of life.

A Breakthrough Moment for Claude Opus 4.5

Yano: I'd like to talk about Claude and Opus 4.5, and I think the best entry point is to mention the "Oh my god, this is over" moment you mentioned on Twitter. There have been several similar moments in the past few years that have given me that same feeling: the release of ChatGPT, the first inference model, and Tesla's FSD V13 (Full Self-Driving). Why did you consider Opus 4.5 an "Oh my god" moment?

Qiao:

I can't explain what happened behind the scenes, but I can tell you how I feel as a user. The last time I wrote code was probably a year ago, but it's been three years since I've seriously written any. From 2010 to 2017, before entering the cryptocurrency space, I worked in quantitative trading, writing code every day, from very low-level C++ to more advanced Python data science code. This was my field for seven years, and I also wrote some front-end and back-end code. But since leaving Messari, I haven't seriously written any code. In recent years, I've only done some personal projects, only finding time to work on some at the end of each year.

Unlike previous versions, Opus 4.5 allows you to quickly assemble a demo, but the last 5% is always difficult for AI to complete completely, still requiring a skilled engineer to handle details such as errors and boundary cases. In Opus 4.5, I only need to tell it what I want to do in simple English, providing a very clear specification. As long as the specification is clear and comprehensive enough, it can complete the task in one go.

Claude and Opus may be going through a hype cycle on Twitter, because I have an engineer friend who told me that OpenAI’s latest GPT-5 Pro is almost as good as Claude.

Yano:

I've used both, but Opus was actually the tool that first got me to switch from ChatGPT. I relied so heavily on ChatGPT's memory feature that I felt I might continue using it because it knew everything about me, but Opus 4.5 is now a better fit for me.

There's a new way of working that I'm unsure whether to categorize as a chatbot or something similar, because the line between the two is blurring. For example, I just did something with Opus 4.5. We have an eight-person sales team at Blockworks, and typically we need to decide who should be responsible for each account and prioritize them, such as Tier 1, Tier 2, etc. Each salesperson has their own judgment, for example, someone might be responsible for 10 Tier 1 accounts, 30 Tier 2 accounts, and 100 Tier 3 accounts. This allocation process is often very complex. So, I delegated this problem to Claude and had it integrate the relevant information. I told it that if a token's FDV (Fully Diluted Valuation) exceeds $1 billion, that account can receive extra priority; if the project associated with that token is a publicly traded company, that's even better.

Claude extracted relevant information from multiple data sources, including CoinGecko and CoinMarketCap, and ranked these accounts. Next, it analyzed the transaction records of each of our team's sales staff with Blockworks over the past eight years to understand the types of products we sold and, based on the likelihood of each sales staff member closing a deal with these accounts, assigned the accounts to the right people.

Qiao:

I also realized what you're describing during my vacation: the line between general chatbots and traditional coding is gradually blurring. When you converse with a general chatbot, it typically generates code on the fly based on your needs. Now, when you need to write code, you don't even have to write it yourself; you can simply express your requirements in simple natural language. This is causing the functionality of code assistants and general chatbots to gradually merge.

Yano: Absolutely right. Then why don't you just use Replit or Lovelace to do your Vibe code work? Why choose Claude?

Qiao:

I think they target different user groups. I haven't tried the latest version of Lovelace yet, but I feel it's more suitable for users who want to quickly build a nice demo or a simple application. It performs well in front-end development, although I haven't actually used it myself. So I initially used Cursor with Opus 4.5, and later tried Claude and Opus 4.5. To be honest, I didn't feel a significant difference in productivity. For me, they performed about the same, but this is just my personal experience.

How is artificial intelligence impacting startups?

Yano: You've witnessed the growth of thousands of crypto startups. How is artificial intelligence impacting these companies now?

Qiao:

The impact of artificial intelligence on startups is enormous, and this change has been particularly pronounced over the past three years. Basically, I ask in every panel, especially the technical partners of each startup, how much have you increased your productivity since ChatGPT launched in 2022? Almost every time, the answer is higher than the last. The trend is remarkably consistent. One recent panel reported a productivity increase of approximately 3 to 4 times.

I believe this impact is more pronounced in early-stage startups and relatively smaller in later-stage large companies because one limitation of code assistants lies in their "context window." For example, suppose Claude can handle a million tokens, meaning it can very efficiently help you launch a new project, and with great results. But if you ask Claude to modify the entire Google codebase, that's virtually impossible, right? Therefore, these kinds of tools are more helpful to small, early-stage startups, while being relatively limited for large companies.

In large companies, the most effective way to utilize code assistants is by establishing clear interdepartmental abstraction layers. This allows complex tasks to be broken down into smaller, more manageable pieces, narrowing the context before being fed into the AI system, making it easier for the system to understand.

Yano:

However, I don't think using the term "productivity" to describe the impact of artificial intelligence is entirely accurate, especially for early-stage startups. If you're a company with 100, 200, or even 1000 employees, a 3x or 4x increase in productivity might be a significant change. But for early-stage startups, they don't even consider the issue of "productivity." Their thinking is more like: Why do we need to hire more people? I often see this in the companies I invest in and in the startups of my friends.

They think that since we have AI, we don't need to hire any new employees at all . This phenomenon is really interesting. For example, I recently created two tools for sales: one is a sales commission calculator. The sales team often asks me how much money they can make. Previously, we needed to ask the finance team to calculate it, but the finance team found it troublesome. Now, with this tool, the problem is solved.

Another tool is a dashboard for sales data. This tool can calculate the cost of the dashboard. Previously, we needed to find a data team and consider whether they used Dune or GoldSky to index the data. If neither, we had to pay extra. Now I've developed a dashboard cost calculator based on all ports.

What's the significance of these gadgets? They mean we won't need to hire a dedicated sales support person anymore, so by 2026, we might see some unicorn startups with only one or two people. These startups may already be taking off, and while they haven't reached a billion-dollar valuation yet, they're already growing rapidly.

Qiao:

Absolutely. I know many people who now run subscription businesses generating $10 million in revenue annually on their own. These people are often former engineers at companies like Meta and Uber who got tired of the red tape at big companies and chose to start their own businesses.

Yano:

However, I've noticed an interesting phenomenon: some of the most successful AI startups aren't companies like ChatGPT or OpenAI, but rather smaller businesses that utilize AI. These companies are reluctant to disclose their core business. Typically, if you have a startup that's doing well and your revenue is growing rapidly, you'd want to tell the world, raise funds, or promote it on social media.

Does the moat still exist?

Yano: What's your view on a company's competitive advantage? Do you think the definition of a competitive advantage will change?

Qiao:

The essence of moats hasn't changed, but moats in the software industry are indeed weakening rapidly. For early-stage startups, there are virtually no moats to speak of. Companies like Facebook, Google, Microsoft, and Apple still have strong moats. AI code assistants cannot destroy these moats. For example, Apple's moat is its developer ecosystem, while Microsoft's moat is the high cost for users to switch from PCs to other platforms. In the cloud services sector, AWS, Azure, and Google GCP's moat is the high cost for customers to switch cloud platforms. Furthermore, there are platforms like YouTube with massive amounts of proprietary data that can be used to develop very powerful video models. Then there are Microsoft's enterprise software, which are critical tools. Of course, you can replicate the functionality of Office, but will enterprises really switch from Office to other software? For them, these tools are too important, and the switching cost is too high.

However, I recently noticed an example that might be one of the biggest undervalued cases in the current financial markets: Adobe . Photoshop and Adobe's creative suite are very well-known, and there's a prevailing view that the latest video and image generation models will replace Adobe's products. But I believe this view is completely wrong because Adobe's competitive advantage lies in its enterprise-level integration capabilities. Many enterprise users of Adobe's creative suite store their images and videos on Adobe's cloud. For these creative professionals, switching from Adobe Cloud to other services is extremely costly.

Furthermore, many creative professionals have been using Photoshop for years, and the operation has become muscle memory for them. Switching to other tools is extremely difficult for them. Therefore, Adobe's current price-to-earnings ratio of only 12 is an incredibly low valuation for such a high-quality company.

Yano: Let's say you work for a company, say a startup with 500 employees that's been operating for ten years. How should the company respond? Many young entrepreneurs might be familiar with these technologies, but for a more established company, it might not be so easy to adapt. Employees might simply dump things into ChatGPT and send out automated emails.

Qiao:

I believe you can't force companies to adopt new technologies. You have to let them find application scenarios where AI technology can have an immediate impact and see the huge impact it can bring. Incidentally, this is actually our experience at Alliance over the past three years. We knew three years ago that AI would become a major trend, and at that time we proposed to become an "AI-first" organization, but we didn't try to impose AI on every corner of the organization.

Instead, we use AI to automate specific processes . For example, we receive tens of thousands of applications every year. Three years ago, I had to personally read all the application documents, about 5,000 a year. This work was extremely time-consuming and exhausting. But now, we've automated about 50% of the work using AI.

Yano: How did you do that? Are you using Opus 4.5? How was this system built? What are the inputs and what are the outputs? Why is it only 50% automated? Why can't it reach 99%?

Qiao:

Yes, this software was developed by our engineers. I believe the code itself is no longer the bottleneck; the key lies in designing appropriate "prompts." I wouldn't call it a secret weapon, because everyone is already aware of some characteristics of excellent founders.

I've simply translated these rules of thumb for successful founders and startups into tips for screening applications. Currently, the system's main function is to filter out clearly unqualified applications, rather than directly telling me who the best candidates are, because I believe this part still requires the judgment of human interviewers. AI hasn't reached that level yet, but I believe that by the end of this year, you'll see AI and AI-driven venture capitalists outperform humans in this area.

Yano: I think the prices of these tools are incredibly low right now, like when Uber first launched, it only cost $5 to travel from New York's Financial District to the Upper East Side. Back then, you'd think that price was obviously wrong, but they were attracting users through venture capital subsidies.

How valuable do you think these tools are? How much do you think they are currently undervalued? For example, how much would you be willing to pay for Opus 4.5?

Qiao:

As for Opus, I'm still not entirely sure; it depends on what I can build with it. But taking Gemini for example, I think its value is underestimated by at least two orders of magnitude. Right now, I only need to pay $20 for the Pro version, and I haven't even upgraded to the Pro Plus yet.

Yano: If Gemini charged $2,000 per month, would you be willing to pay it?

Qiao:

Yes, because it's incredibly powerful. In fact, as I mentioned on Twitter, it's a research assistant, a junior researcher, a junior code assistant, a pretty good medical advisor (can verify doctor's advice), and a pretty good legal assistant. All things considered, if you put all those features together, the $2000 price tag is an absolute steal.

The decision to invest in Google

Yano: How much time do you spend using apps like Gemini and Claude each day? What daily activities do you squeeze in to make time for them? For example, do you reduce your Zoom meeting time, or spend less time on social media?

Qiao:

Actually, I analyzed my time allocation six months ago, which prompted my decision to invest in Google. I checked my iPhone usage history and found that my three most used apps were Chrome, YouTube, and Gemini , all of which are Google products.

Of course, I was initially worried that ChatGPT would threaten Google's search business, so I talked to my wife about it. She gave me some insights, saying that she uses Google search most often for shopping, and ChatGPT cannot completely replace this function in the short term. Then I did some research and found that more than half of Google's search revenue actually comes from shopping ads, which made me feel that Google's core business is still very solid.

No one can shake its position in the short term. Furthermore, Google has technological advantages such as GCP (Google Cloud Platform) and TPU (Tensor Processing Unit), which I only realized later. I believe these factors all indicate that Google's competitive advantage is extremely strong. Therefore, I decided to invest in Google last year, which was almost my only major investment that year.

Yano: Last Lunar New Year's Eve, I had dinner with a Google employee, and we talked about these things. Everyone was discussing their favorite tools. The Google employee mentioned that many people overlook the fact that Google possesses massive amounts of shopping data, and the market doesn't seem to fully realize this. So, are there other similar opportunities?

Qiao:

I think Adobe might be the Google of this year because its current valuation is very cheap, and the market's view of it is somewhat similar to the view of Google back then. Actually, I've never used it myself, but I know it has many use cases.

Yano: No new users will use Adobe. Like Google, my go-to app is YouTube; I use YouTube very frequently, and I also use Gemini. But Adobe? I don't use it at all, so I don't think new users will choose Adobe either.

Qiao:

But all new users will use Google, and I think a common misconception in the market is that Adobe is not a consumer-facing product, but an enterprise-facing one.

Yano: Yes, it's an enterprise-level product, but my company would never consider using Adobe; Figma is much more popular.

Qiao:

Figma and Adobe actually serve two different markets. Adobe has lost out in the market competition for new project demonstrations and website development, where Figma dominates. What about Canva?

Yano:

Canva targets the low-end market and is more suitable for casual users, while Adobe serves high-end enterprise users. Therefore, I agree with your point that Adobe's new user growth has indeed stagnated for a long time, but its pricing power is very strong, allowing it to continuously increase subscription fees.

How to allocate time reasonably in 2026

Yano: As a Google user, I'd like to talk about investing, but I'd also like to explore a question: How should we allocate our time in 2026? How do you think artificial intelligence will change the way we allocate our time?

Qiao:

Yes, I believe everyone should learn to code, and must learn. Otherwise, they risk falling far behind. However, this doesn't refer to traditional programming, but rather to using automation tools to optimize certain aspects of one's life and work . I believe that in the future, there will be many B2B SaaS software tools, such as Gmail and Zoom, and people will still pay for these general-purpose tools. At the same time, everyone's workflow will have specific, highly personalized needs, and third-party software companies may not develop tailored solutions for these needs.

Yano: For example, we can develop our own commission calculator at Blockworks. This is a good example; nobody would develop a dedicated SaaS software to do this kind of thing.

Qiao:

Furthermore, there's no need to be alarmed when I say "everyone should learn to code." Modern technology has already eliminated the need for writing complex code. Automation can be achieved simply by communicating with systems using natural language.

Therefore, I strongly recommend that you don't necessarily need to use complex tools like Claude Opus. I can recommend some simpler, easier-to-use tools. For example, Replit. I first realized the potential of artificial intelligence through Replit. That's when I understood that AI is not just a chatbot; it can help you build anything. This technology will completely change our world. I highly recommend everyone try Replit; it's really cool.

Yano: What impact do you think this gap will have on the labor market?

Qiao:

I think this will be similar to the widespread adoption of the internet ; these tools will make already efficient and capable people even more efficient and intelligent, while those who are inefficient may fall further behind. Artificial intelligence is an extremely powerful tool that can significantly improve efficiency, but ultimately it depends on how much you are willing to use it.

Investing using AI models

Yano: In a sense, from the opposing perspective, do you think the Warren Buffett stock tracker you spent time building was really worth it?

Qiao:

Yes, it did take me quite a bit of time. While the result was good, more importantly, I gained a deeper understanding of the relevant technologies in the process of building this tool.

This tool can be seen as a digital clone of investment masters such as Warren Buffett, Charlie Munger, Howard Marks, Peter Drucker, and Bill Miller. It regularly scans thousands of stock codes and conducts in-depth research on each one, attempting to simulate the investment logic of Buffett and Munger.

The code itself is quite simple, but designing the prompts took me several months of repeated adjustments. I designed a very detailed set of prompts to simulate the thought process of Buffett and Munger in evaluating potential investment opportunities.

This process consists of six steps. First, it uses a deep research model to collect information on six key aspects that they might be interested in. Then, it calls another API to use an inference model for analysis. Deep research and inference modeling are two separate steps.

Deep research models excel at collecting facts and data, while inference models, although sometimes generating erroneous information, perform better in logical reasoning. Once accurate data is fed into an inference model, its analytical capabilities far surpass those of a deep research model.

Next, in the second stage, I will invoke the inference model to simulate the investment committee of a "digital Buffett and Munger" to analyze whether it is worthwhile to invest in a particular stock. Ultimately, it will output a specific recommendation, and that's the whole process.

Have you heard the theory that Renaissance Technologies discovered Large Language Models (LLMs) long before anyone else, but kept the technology secret, which might explain their incredibly high returns on investment? However, my initial design of the Buffett and Munger models was precisely to avoid competing with Renaissance Technologies. They excel at short-term trading, such as day trading or week-long operations. On such short timescales, the latest AI models truly struggle to compete.

Yano: So are you also competing with companies like Susquehanna?

Qiao:

That's right, but my model is more geared towards long-term investing. In today's market, almost no one has the patience to hold a stock for more than 5 minutes, which is where language models can come in.

Yano: So, how do you combine the advice of different investment masters? For example, the advice of Howard Marks and Peter Drucker may be completely different, and their thinking is completely opposite to that of Buffett.

Qiao: I will take their suggestions into weight and average them.

Yano: They all recommend buying Adobe?

Qiao:

Yes, another interesting phenomenon is that if you run the same prompt multiple times, the model may give different answers each time. So if you run the same question multiple times and average the results, and each result suggests buying a certain stock, then the credibility of that suggestion will be greatly improved.

It recommended about ten stocks, four of which are already in Berkshire Hathaway's portfolio. These include Chubb, an insurance company, and Google.

How is artificial intelligence changing branding and distribution methods?

Yano: How will artificial intelligence change the way brands communicate and distribute content? For example, Delphi, which we mentioned earlier, and I wonder if you've noticed OpenAI's recently launched Sora tool. With Sora, you can clone someone's image and embed it into your videos. This reminds me of the changes in marketing and branding. In the near future, brand marketing will clearly become more personalized.

Qiao:

This trend has actually already begun. A few days ago I saw an ad on LinkedIn that said, "Hey Jason, as the co-founder of Blockworks, I really think you'll love Rippling."

Yano:

Artificial intelligence will make their advertising algorithms more intelligent, and ads will become highly precise. Actually, I searched for the term "High Rocks" online a few days ago. It's a marathon-style fitness activity similar to CrossFit. I've been training for High Rocks recently, so I searched for High Rocks fitness apps on the App Store and found several specifically for High Rocks training. But High Rocks is a very niche activity; logically, there shouldn't be so many dedicated apps.

I found this strange, so I wondered if these apps were actually designed for High Rocks, or if they were simply tailoring ads to users searching for "High Rocks." Sure enough, these were just regular fitness and nutrition apps, but their advertising was incredibly targeted, so I think this kind of personalized advertising will become more common in the future.

Health and Longevity

Yano: What do you think is the most effective change in terms of health?

Qiao:

Ultimately, the most important factors are still three basic ones: diet, sleep, and exercise . Back in 2021, I was very focused on optimizing details, such as trying various health supplements and saunas. But after four or five years of personal practice, reading a lot of research, and listening to countless podcasts, I've found that nothing is more effective than getting eight hours of sleep every day, maintaining a healthy diet, and sticking to a regular exercise routine.

Yano: So how do you plan your diet and exercise?

Qiao:

To be honest, I no longer strive for extreme dietary optimization. Doing so puts too much pressure on me. I simply try to maintain a healthy diet and not burden myself with too much stress.

Yano: This reminds me of a neurological meme where the person in the middle does complicated food preparation every day and takes 17 kinds of supplements every morning. Your current attitude is "Just eat healthily, no need to overthink it."

Qiao:

Yes, I find that if I try to optimize everything to the extreme, like Brian Johnson (note: Brian Johnson is an entrepreneur known for his extreme health management), I experience a lot of stress. And stress leads to elevated cortisol levels, which is not good for longevity.