#美国与科技巨头投18亿美元建ai生物学数据集
👉 AI基建怎么看，币安聊天室跟进
The U.S. government and several tech giants pooled some money.
The amount is $1.8 billion.
The purpose is to build AI bioinformatics datasets.
One of the funders is also a nonprofit research organization.
Where does the weight of this matter lie?
AI’s bottleneck is shifting from compute to data.
Data is being treated as a new means of production.
Publicly available text is running out.
High-quality professional data is becoming increasingly expensive.
In fields like biology and medicine, data is even harder to obtain.
Whoever builds the library first will have a better-stocked model first.
For the industry, this is like building roads.
Datasets are public infrastructure.
Once built, the applications on top can finally run.
Drug discovery, diagnostics, materials—all are waiting for this batch of data.
Early investment is high, and the payback cycle is long.
There are challenges too.
How do we determine data privacy and ownership?
Standards and formats still need to be unified.
Even if the money is in place, data providers may not be willing to come.
For encryption, it’s about data title/ownership and transactions.
It’s the next piece of land that the blockchain is competing for.
Do you think data will become AI’s new bottleneck? Let’s discuss in the comments.
👉 AI基建怎么看，币安聊天室跟进
The U.S. government and several tech giants pooled some money.
The amount is $1.8 billion.
The purpose is to build AI bioinformatics datasets.
One of the funders is also a nonprofit research organization.
Where does the weight of this matter lie?
AI’s bottleneck is shifting from compute to data.
Data is being treated as a new means of production.
Publicly available text is running out.
High-quality professional data is becoming increasingly expensive.
In fields like biology and medicine, data is even harder to obtain.
Whoever builds the library first will have a better-stocked model first.
For the industry, this is like building roads.
Datasets are public infrastructure.
Once built, the applications on top can finally run.
Drug discovery, diagnostics, materials—all are waiting for this batch of data.
Early investment is high, and the payback cycle is long.
There are challenges too.
How do we determine data privacy and ownership?
Standards and formats still need to be unified.
Even if the money is in place, data providers may not be willing to come.
For encryption, it’s about data title/ownership and transactions.
It’s the next piece of land that the blockchain is competing for.
Do you think data will become AI’s new bottleneck? Let’s discuss in the comments.