sysfoc
Home / News / high-ai-agent-failure-rate-generally-for-web3
high-ai-agent-failure-rate-generally-for-web3
2025-01-28 11:28:56

Experts: High AI Agent Failure Rate Generally for Web3, but ‘Not Necessarily a Bad Thing’

Jenny Jones-author-image Jennifer Michele
50 views

Given that AI agents and their tokens decrease practical uses, apparent utility, and measurable value, it is not surprising that there is an alarmingly high number of failed AI agents. According to some experts, even more AI agents are likely to drop in the future, much like startups, which have a projected survival rate of less than 1%.

Almost 90% of AI agents launched between January 2 and January 14, 2025, have failed. Furthermore, nearly 75% of traders in AI agent tokens—which have an average lifespan of just 17 days—are falling money. These results show the significant risks associated with trading AI agent tokens.

AI Agents' Failures Explained

Michael Heinrich, managing director of 0G Labs, claims the number of defunct AI agents shouldn't be a reason for alarm, even though the high failure rate also indicates Web3's extraordinarily short attention spans.

This is actually motivating for the industry and isn't always a bad thing. It indicates that only AI agents and projects with true worth and utility will endure. Heinrich revealed that this type of natural selection is beneficial for the space's long-term health.

Hence, founders and entrepreneurs serving with AI agents should view the failure rate as proof of the "huge perceived demand" for AI agents rather than as a cause for pessimism. However, at the moment, the vast demand is surpassing the capacity to offer valuable and practical solutions.

Differentiating Real AI Agent News from Hype

Meanwhile, DWF Labs colleague Lingling Jiang recommended against falling for too optimistic AI agents. Jiang asserted that "there are no shortcuts to creating reliability or longevity." "The key to the success of your project is strategic vision, hard work, and collaborating with wonderful, like-minded partners."

It is nevertheless expected that AI agents will become increasingly recognized in 2025 despite their seeming dismal success record. As a result, more users will probably be encouraged to interact with them.

Renç Korzay, CEO of Giza, suggested that consumers experimenting with AI agent tokens focus more on execution and less on marketing narratives while making judgments to decrease the likelihood that they will experience the same end as their trailblazing peers. "Users should request evidence of real utility, not just token economics or technical whitepapers," Korzay stated.

Heinrich suggested that potential investors in AI agent tokens should use online monitoring tools such as Dune Analytics or Token Terminal to get more in-depth information. Jiang claimed that it is crucial to stay up to date on events and developments and separate fact from fiction.

Shashank Yadav, CEO and founder of Fraction AI, claimed that smart contract lockups with linear vesting are necessary because of the high failure rate. Yadav stated, "Mandatory TEEs for tamper-proof agent works and zero-knowledge proofs for transparent yet competitive execution" are also essential for this.

The experts almost unanimously believed that community feedback is crucial to a project's success. Community feedback ensures that AI agents develop to "solve real problems and provide genuine value," from helping to shape and promote the agents to finding problems before tokens perish.

Most Read

Lorem ipsum dolor sit amet consectetur adipisicing elit. Suscipit!

Comments

Want to add a comment?