Author: TVBee
Both stock market and crypto players are worried that AI might repeat the 2000 internet bubble, but in reality, Web3 is the true replay of the internet bubble.
Bubble: Supply-Side Self-Indulgence, Lackluster Demand
During the 2000 internet bubble phase, capital was self-indulgent on the supply side. Many internet-listed companies had no profits or cash flow, yet their stock prices soared wildly.
On the other end, internet applications on the demand side were pitifully few... Before 2000, there was no Douyin or Kuaishou, no Alibaba or JD.com, no WeChat.
The mainstream internet applications at the time were only the雏形 of QQ like OICQ, MSN and other chat software, download tools like FlashGet and NetAnts, information platforms like Yahoo,网易, and搜狐, basic applications like Google Search and email, and the early stages of e-commerce like Amazon and eBay, which had very few users.
At that time, there was no mobile internet. Personal computers were considered luxury items, and most households in China did not have one. Monitors were those big, bulky CRT monitors. Laptops were as thick as bricks, and internet access relied on dial-up connections through telephone lines. Even older were the storage media; the primary storage was floppy disks—yes, the A drive. Their capacity was only 1.44MB, not even enough to hold a single moderately sized image today.
It's worth mentioning that there were single-player games back then, but they were pixelated games.
AI: Infrastructure Leaders Soar, Products Growing
Looking at the AI industry, stock prices have surged, but the main gainers are AI leaders, primarily in infrastructure—NVIDIA, AMD, SK Hynix, Samsung, Micron... This is because AI requires extensive training before deployment. On the supply side, AI does not have the exaggerated bubble seen in 2000. At least these leading AI companies do have considerable profits.
On the demand side, we can at least see Gemini, Claude, GPT, Doubao... various UGC applications. And the recently popular Lobster, the newly launched Perplexity Personal Computer... the highly eye-catching AI robots at the Spring Festival Gala... Although the breadth of AI products still can't be compared to that of the internet in 2000, this is because AI development has high infrastructure requirements. AI needs to develop its infrastructure first before it can integrate into more types of applications.
Web3: Technology as a Gimmick, Applications Few and Far Between
Circulating Market Cap and Profit
Now look at the Web3 industry. Various technological narratives emerge one after another, but how many applications actually have users? Aside from a handful of DeFi leaders, there might only be MEME platforms, prediction markets, and Perp DEXs, with much of the activity in the latter two coming from interaction behaviors driven by airdrop expectations.
Corresponding to the scarcity of products on the demand side is the self-indulgence on the supply side.
⏵ Example from the ZK track: ZKsync with a daily income of $458 has a circulating market cap of $176 million, equivalent to a P/E ratio of 1052.
⏵ Example from the L2 track: Optimism with a daily income of $2,427 has a circulating market cap of $253 million, equivalent to a P/E ratio of 285.
⏵ Example from the L1 track: Sei with a daily income of $3,564 has a circulating market cap of $424 million, equivalent to a P/E ratio of 327.
And there's Blast with a daily income of -$6...
If we follow stock market logic, shareholders of the first three projects would have to wait 1052 years, 285 years, and 327 years respectively to break even, and this doesn't even include the infrastructure and operational costs required to keep these chains running.
Although these ecosystems are not meant to break even and profit from on-chain revenue, for shareholders, i.e., token holders, this is a nightmare...
What Web3 Applications Are There?
Now, what Web3 applications are there?
Former applications like the metaverse,链游, inscriptions, social... and probably some others we can't even remember are almost never mentioned now.
Currently, besides DeFi and RWA, the only ones with users are almost just meme coins, prediction markets, and perp DEXs. Among these, meme coins are PvP among existing capital, and part of the activity in prediction markets and Perp DEXs is still driven by interaction behaviors due to airdrop expectations.
Compared to the internet applications mentioned from 2000, Web3 applications are truly few and far between...
Final Thoughts
Therefore, the sector that is truly indulging on the supply side while lacking achievement on the demand side is not AI, but Web3.
US stock market players overlap with those from 26 years ago. Investors and Wall Street are deliberately avoiding a repeat of history. AI has a bubble, but it differs from the 2000 internet bubble.
Instead, it's the younger Web3 field where capital is speculating on technology on the supply side, while the demand side has few truly practical products. This is the true reappearance of the 2000 internet bubble.
Based on the reasoning of a moderate AI泡沫 and a high Web3泡沫:
- First, the US stock market will correct, but a crash like in 2000 is unlikely.
- Second, BTC, which is correlated with US stocks, will be moderately affected.
- Third, just like the 2000 internet bubble,山寨币 will continue to be washed out, forming a painful process of separating the wheat from the chaff. This process might be even longer than I imagine.
So far,山寨币 have generally been falling for about 15 months since the end of 2024, but this is not the end.
Don't believe the talk that山寨币 have bottomed out. Some have, others have no bottom.
