The thing that makes Devin "the AI coder" really interesting is the new modality of interaction.
@emollick42万 位关注者
75,347 次观看
9 条评论
评论来自 X (Twitter) 上的原帖

@emollick
To be clear, AI systems aren't quite there yet to do this work autonomously and error-free without help. Even afterward, there is a way to go before you would want to trust a major project to AI, but it is a fascinating start nonetheless.
44 个赞

@emollick
I can't code in Python, it is doing everything autonomously based on my request for a "website that explains how dilution works in startup funding."
33 个赞

@yoheinakajima
Agreed! I think tasks should run in parallel to the chat, which is what I tested with babyfox Feels much faster when you can ask for the next task before the last task is done.
10 个赞

@RDPandole
The LLM market is saturated as we have used up almost all of the text data publicly available on the web to train these models. So how do we scale these models from here in order to make them even more intelligent? Vision is the obvious answer. Multi-modal models are the future; however, the underlying architecture for these vision models will be different from the autoregressive models used by the LLMs.
7 个赞

@enginyer
This looks like the best UX for AI coding so far. I want to test it now!!
3 个赞

@singular_prism
How long before OpenAI copies this
2 个赞

@daniel_mac8
Can't wait to get my hands on this
1 个赞

@matt_barrie
How much is it costing per hour
1 个赞

@DaniAcostaAI
Awesome. How did you got access?
1 个赞