2026-01-19 19:37:00

Web2 companies are accelerating their shift towards synthetic data, which is a major trend. Botanika founder Siwon Kim mentioned in a recent tech talk that the logic behind this shift is very clear: first, significantly reduce compliance costs; second, eliminate the risk of sensitive data leaks; third, AI model training no longer relies on massive raw data. This is not a future scenario but happening now. From the enterprise perspective, synthetic data addresses pain points in traditional data processing—protecting privacy, improving efficiency, and accelerating the deployment of AI applications. It is foreseeable that this paradigm shift in technology will further drive the development of the entire industry.

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

14 Likes

Reward
14
7
Repost
Share

Comment

0/400

HappyToBeDumped

· 16h ago

Synthetic data should have been popularized long ago; it's really convenient, and companies have finally realized it.

View OriginalReply0

SerLiquidated

· 17h ago

Synthetic data is indeed gaining momentum, and privacy cost issues are a real pain point.

View OriginalReply0

AirdropBlackHole

· 17h ago

Synthetic data is really unavoidable now; companies have long been fed up with privacy compliance. Saving costs and avoiding risks—who wouldn't want that?

View OriginalReply0

MemeKingNFT

· 17h ago

Synthetic data has been overdue. Web2 is still using outdated data processing methods, while we on the chain have already innovated. Lower costs, prevent leaks, faster training—basically the idea of "decentralization" of data. It's just that they are only now waking up to it. I’m a bit skeptical about their reaction speed.

View OriginalReply0

RektCoaster

· 17h ago

Synthetic data has been overdue; the previous method of stacking raw data was simply too inefficient. But how many companies can truly implement it?

View OriginalReply0

WhaleStalker

· 17h ago

Synthetic data should have been popularized long ago. Companies are overwhelmed by privacy leaks and related issues, and now that solutions are available, they're still hesitating?

View OriginalReply0

BackrowObserver

· 17h ago

Synthetic data is indeed on the rise, but only a few companies can truly utilize it effectively. Synthetic data is popular, but how can data quality be guaranteed? Siwon Kim is right, saving costs and reducing risks, but who will bear the problem of model hallucination? In this wave of AI, synthetic data might be the next overhyped concept. Honestly, isn't privacy protection and efficiency improvement inherently contradictory? Another "revolutionary technology," let's see what it looks like in five years. Lower compliance costs sound great, but the excuses for cutting corners are also increasing.

View OriginalReply0