Crewai Raises $18M–But Are AI Agents Ready for Prime Time?
finance.yahoo.com
finance.yahoo.com
From what I’ve seen, AI agents aren’t even close to being reliable for real-world tasks yet. What am I missing here? Is there something about Crewai’s approach that’s genuinely groundbreaking, or are they just riding the AI hype?
Would love to get educated on this. Links:
https://www.lycee.ai/blog/ai-agents-automation-eng Crewai’s funding announcement: https://finance.yahoo.com/news/crewai-launches-multi-agentic... Looking forward to hearing your thoughts!
The founder seemed super enthusiastic and constantly tweeting about how great everything was. My guess is it's mostly hype but I guess now they have funding they might prove me wrong.
My first conclusion is that it’s mostly hype. And honestly, $18M isn’t nearly enough to make AI agents reliable and useful, because solving that would require breakthroughs in reasoning—something we're nowhere close to. Even o1 fails miserably at most real life reasoning tasks despite the hype from openai. (https://www.lycee.ai/blog/why-sam-altman-is-wrong)
It seems Crewai is more focused on deploying AI agents now, hoping that reliability will eventually come from improvements made by foundation model companies. But that's a highly uncertain, almost wishful, scenario. And they shouldn't go around screaming that AI agents are reliable in production yet.