Silicon Valley Watch: Finding Certainty in the AI Wave

Author: ChichiHong, Co-Founder of ScalingX Labs

Between the hills and sea fog of San Francisco, AI is rewriting the rhythm of the Bay Area at a visibly rapid pace. For ChichiHong—ScalingX Labs co-founder, deeply rooted in Web3 and now in North America—the strongest feeling isn’t that one place is running far ahead of the rest, but rather: the Bay Area is forming a “multi-point bloom” pattern, made up of San Francisco, the South Bay, and surrounding cities.

In her daily movement, San Francisco brings together large model and AI infrastructure companies, the South Bay still carries traditional tech giants and engineering communities, and nodes like Palo Alto are filled with Demo Days of all sizes, incubators, and startup events. When everything is accelerating, replacing, and rearranging, what she keeps thinking about isn’t “where is the center,” but rather: in such a multi-center AI wave, what relatively certain things can people still hold on to—whether it’s geographic choices, track judgments, entrepreneurial paths, or their own lives and mindsets.

1. Geographic Choices: Multiple Growth Points

In recent years, San Francisco has been reshaped into one of the densest stages for generative AI companies, thanks to the headquarters and expansion of major model companies such as OpenAI and Anthropic; new stories, new companies, and new AI narratives mostly originate from here.

At the same time, the South Bay remains the base for large tech companies such as Google and Meta, as well as numerous chip and cloud infrastructure enterprises. It gathers a huge number of experienced engineers and foundational technology teams, continuously attracting and exporting talent around the world.

In the stories she hears, two scenes often appear side by side: some people sell their companies and later buy million-dollar homes in San Francisco, betting their future on AI and new wealth narratives; others, even though their large employers are laying off staff, are quickly poached by other teams or startups, and the South Bay’s housing prices and community atmosphere haven’t noticeably cooled because “AI has taken the spotlight.”

For her, this state of “both the new and the old are growing” is itself a form of geographic certainty:

San Francisco represents new stories, new companies, and new opportunities—it’s the most concentrated stage for AI narratives;

The South Bay represents the old system, mature engineers, and stable infrastructure—it continues to attract and deliver talent;

Neither side has losers—only the roles they play are different.

In such a landscape, the question is no longer “should you leave the South Bay and move to San Francisco,” but a more fine-grained choice: which type of resources do you need to be closer to—new tech companies and capital networks, or established tech giants and the engineering ecosystem. For those who want to stand firm in the AI wave, this reality of “the new and the old are both booming” instead provides a kind of predictable geographic security: no matter which side you stand on, there are people and matters worth connecting with.

For her, the first layer of “certainty” is already quite clear:

The geographic focus is concentrating toward San Francisco;

The South Bay still carries large employers and existing engineers, but the balance of voice and imagination is moving northward.

For entrepreneurs and investors who want to get close to the AI frontier, “being in San Francisco” itself is already the most straightforward geographic certainty.

2. Track Choices: AI and Web3

Chichi, who comes from a Web3 accelerator, inevitably gets asked: in combining AI and Web3, is there truly a new and sufficiently certain direction? Her answer differs from many optimistic narratives—in the past year, she hasn’t seen any new path that could be called a “paradigm shift,” and most so-called “AI+Web3” projects are still using stories that were already being told last year.

In her view, the most honest judgment right now is:

AI’s certainty is much stronger than Web3’s. Nearly every industry is actively looking for ways to make use of AI—from development and marketing to customer service—AI has already become infrastructure;

Web3’s demand for AI is clear. On-chain projects need AI for automated operations, content production, and user reach; even in risk control and data analysis, AI also has clear advantages;

AI currently has no urgent need for Web3. To prove that “without blockchain, AI can’t run,” there are not yet enough convincing cases.

She sums up this asymmetric relationship with a memorable line: “Everyone needs AI, Web3 also needs AI, but AI doesn’t need Web3.”

This doesn’t mean Crypto is being completely sidelined. Over a longer cycle, many US local investors still believe that the risk-reward ratio of crypto assets may not be worse than any single AI track. What’s truly worth pondering is that stablecoins have quietly entered the “back-end system” of AI.

According to Circle’s data, over the past 9 months, about 400,000 AI agents completed 140 million payments totaling 43 million US dollars. Of that, 98.6% was settled through USDC, and the average amount per transaction is only 0.31 US dollars—meaning micro-transactions between machines are already continuously happening in a crypto-native way. In this sense, some people working in the AI sector aren’t just verbally “believing in Crypto”; they are treating stablecoins as the default payment layer for agents, effectively connecting the two tracks at the behavioral level.

It’s just that, at this point in time, if Chichi is talking about “certainty in tracks,” she still prefers to see AI as the foundation for all industries, and to see Web3/stablecoins as extremely suitable “infrastructure plugins” in certain scenarios—rather than forcibly tying the two pieces together and using a composite narrative to explain all problems.

3. Certainty in Entrepreneurial Paths: Small Teams vs. VC—Not an Either/Or

The impact of AI on entrepreneurial paths, Chichi summarizes as “threshold restructuring.”

What left the deepest impression on her was a recently viral Medvi case—a remote medical services company built around the weight-loss drug GLP‑1. The founder, Matthew Gallagher, had an ordinary background and wasn’t a top-school prodigy. From his home in Los Angeles, he used about 20,000 US dollars and a dozen AI tools, spending two months to build up everything step by step: the website, the appointment process, the consultation questionnaires, ad creatives, and customer service replies.

The emergence of these “one-person companies” or “a few-person companies” brings new certainty to entrepreneurial paths:

What can be confirmed is: with AI properly leveraged, the ceiling for small teams is greatly raised, and starting a business no longer necessarily means first assembling a team of a dozen or so;

What can also be confirmed is: not all projects, because of this, “no longer need VC.”

Chichi emphasizes that she sees two realities existing at the same time:

On one side, there are more and more cases of “a great company can be built without funding”—revenue can be generated with just a few tens of thousands of dollars, sustained by rolling, continuous growth, without necessarily following the traditional funding tempo;

On the other side, there are directions that truly require heavy resources and heavy investment: computing power, hardware, complex infrastructure, and strong compliance scenarios. For these projects, without VC funding and resources, it’s hard to get in during the window period.

This directly changes her understanding of “VC certainty.” In the past, it might have been “money first, then product,” but now it feels more like:

Truly excellent entrepreneurs who know how to use AI are less dependent on money early on, and don’t need to compromise too much in order to “get off the ground”;

If VCs want to maintain their own certainty, they must shift from “providing money” to “providing resources,” such as GPUs, talent networks, channels, and brand endorsement.

She describes Silicon Valley as: “Demo Days happen almost every day.” Various incubators and event spaces provide almost unlimited opportunities for founders and investors to connect; investors can even leave direct comments under X or Product Hunt saying, “Want to invest in you,” and some funds deliberately seek to place early bets on “high school geniuses.”

In such an extremely active, highly disintermediated fundraising environment, her advice to founders is:

Don’t rush to treat “whether to raise funding or not” as a binary question;

First use AI to get the product running, and then determine whether you need “money,” or whether you need “resources + brand + ecosystem”;

Treat VC as an amplifier, not a starting point.

4. Conclusion: In Uncertainty, People Always Learn How to Adjust Themselves

Amid technology and development that are becoming increasingly exciting, Chichi sees the same force reflected across different interfaces: AI is rewriting existing order at a very high speed—company maps are shifting, track boundaries are blurring, entrepreneurial paths are being compressed, and the relationship between people and the world is also being renegotiated.

The more hidden layer has nothing to do with cities or valuations. The people she met in HK and Silicon Valley—middle-aged finance practitioners who worry that if they can’t keep up with AI, they’re done; engineers at large firms repeatedly hit by layoff emails and visa dates being pressed over and over—made her realize that insecurity has become the baseline noise for people today. It won’t disappear just because you’re in a big company or how many stocks you hold; instead, in an environment with higher information density and faster rhythms, it’s continually amplified.

Therefore, “seeking certainty in the AI wave” ultimately can hardly stay at the level of discussions about cities, tracks, or capital. Inevitably, it falls back into a more personal dimension: in such an environment, whether people are still willing—and still dare—to proactively adjust themselves.