#FIL# Messari: When you store data on the Filecoin network, you can actually prove and verify the source of the data.

As tech giants compete by making large-scale investments in artificial intelligence (AI) infrastructure, Web3 companies are developing an alternative, decentralized model.
According to data from the Dell'Oro Group, large tech companies (Amazon, Google, Meta, and Microsoft) are expected to invest approximately $180 billion in artificial intelligence infrastructure and data centers in 2024, with spending projected to reach $320 billion in 2025.
The founder of the Filecoin Foundation, Clara Tsao (, stated: "It's always difficult to fight against the giant Goliath, but we are like the little David."
During the proxy day at EthCC, the Incrypted team spoke with her, Aethir CEO Dan Wang, and Messari's head of partnership Brandon Slack about how Web3 companies are building the future of AI and the challenges they face in the process.
Current status of infrastructure: complexity and cost
Cao pointed out that building artificial intelligence systems is far more expensive and complex than imagined. According to her, today's systems require multiple components: data layer, computing layer, GPU, and AI agents.
"One problem we have noticed, which may not only be a pain point in the decentralized artificial intelligence field but also one of the biggest pain points in the entire artificial intelligence service sector, is listing, listing time, and scalability. I believe people underestimate the cost of operating artificial intelligence services on a large scale," Wang from Aethir added.
He said that once the project starts to scale, the initial costs will skyrocket. The CEO of Aethir pointed out that this is why most AI agents cannot enter the mass market—they simply lack a suitable profit model.
As mentioned earlier, large technology companies are preparing to invest hundreds of billions of dollars in AI infrastructure. In addition, they have other advantages in the Web3 space.
"There are many challenges, not just for us, but for the entire Web3 ecosystem, as we are up against traditional monopolistic companies. Our biggest competitors are clearly traditional cloud storage, which has significant advantages in terms of time span and user base," Clara Tsao emphasized.
It is worth noting that, according to Messari, the market value of decentralized physical infrastructure networks (DePIN) has surpassed $40 billion, and Filecoin is one of the pioneers in this field. Analysts from the World Economic Forum predict that by 2028, this figure could rise to $3.5 trillion.
Trust but verify
Besides cost, a crucial factor in using artificial intelligence and data is their accuracy and quality.
Brandon Slack from Messari stated, "The biggest problem in life, especially in the fields of data and artificial intelligence, is GIGO: Garbage In, Garbage Out, because everything is advancing too quickly and there is so much information to sift through."
This is precisely the advantage of a decentralized approach. In traditional cloud systems, users often cannot access verified sources of information or verify the authenticity and immutability of the data.
In an era of deep forgery and the proliferation of false information, this is crucial. Nowadays, there is an excess of data on the internet, and people often believe everything easily—therefore, verification is essential.
The ability to prove the source of data is a fundamental advantage of Web3: blockchain not only stores information but also ensures its authenticity.
"We are building many features to integrate and store data from agents to agents. You can imagine that AWS or centralized storage facilities may have considered this, and may not have... Being able to prove the source of the data is the real solution to this problem. When you store data on the Filecoin network, you can actually prove and verify the source of the data." — Clara Tsao said.
Specialized strategy
Another advantage of decentralization lies in the specialization focused on specific areas, while large tech companies attempt to solve "all problems that everyone can solve." Slack likened this to medicine—general artificial intelligence is like a therapist, but there are also experts in specific fields. In his view, the development of artificial intelligence will follow this path:
"There will be a Master of Laws (LLM) program, which is great for retail or simple issues. ChatGPT is a tool I use every day. But if you want to get the true and accurate source of information about cryptocurrencies and really filter out the noise, I think tools like Messari Copilot can help you find answers quickly."
Innovation at the Intersection of Web3 and AI
New technologies are particularly active in the financial sector.
Artificial intelligence agents on the blockchain have reached an impressive scale. According to VanEck, this sector has the potential to create over one million new agents by 2025.
Clara Tsao stated: "We have seen the rapid growth of AI agents in the Web3 space, and many people are now using them for trading, especially in the DeFi sector."
Data storage remains a key link in this process. Cao explained that:
We are very much looking forward to the future significance of Filecoin and large-scale data storage in the current field of artificial intelligence. Most AI agents, whether from centralized or decentralized systems, rely on the amount of training data to ensure accuracy. Storage will never disappear.
Can Web3 projects change the market?
The artificial intelligence revolution has only just begun. Web3 companies are betting on decentralization and specialization, in stark contrast to centralized monopolies. In addition, blockchain provides transparency of data sources.
History shows that in the battle between David and Goliath, it is often the quick and focused newcomers who achieve victory, as they can find moves that even the big companies do not think of.