AI's first real victims are questions and answers websites, not Google
The actual battle is happening between ChatGPT, Anthropic, Bard etc. and the Q/A sites like Quora and StackOverflow.
Mass media and social networks are entirely focused on the danger of OpenAI to Search Engines like Google, but the actual victims of the AI battle are rarely mentioned. Questions and answers sites like Quora and StackOverflow are the first dominos to fall in the GPT disruption wave.
When you think about it for a second, it’s evident that the user interaction model on Q/A sites is very similar to the interaction through a conversational interface like ChatGPT.
Asking a question on Quora or Stackoverflow often fetches a large number of incorrect or poor answers and a small number of high-quality answers, making it an ideal replacement candidate for chatbots like ChatGPT.
ChatGPT is not completely accurate in its answers, but most Quora answers are not either, so the users might very well use the simpler alternative of a chatbot.
If ChatGPT prevails in this battle, Quora could lose hundreds of millions of page views per month.
The danger is real. So what is Quora doing to prevent its demise?
StackOverflow decided to fight against ChatGPT by banning GPT-generated content, Quora decided to join forces with it and build a new product.
They just recently launched its own chatbot product called Poe - www.poe.com
Poe is the most advanced chatbot interface to date, and it aggregates several chatbots (including ChatGPT and Anthropic) into a single interface as individual bots, with fancy names like “Heron” and “Dragonfly”.
You can interact with each bot separately, almost as if they are separate users on the platform.
I think that Poe gives us a glimpse of where the whole AI space is going. Aggregating multiple bots in an interface, giving each bot an identity/profile, and then integrating them as “smart users” into the user interfaces.
In the future, we might see a “Dragonfly” user answering questions on the main Quora interface along with human users, and then humans will fight against AI to provide a more detailed and authentic answer on each Q/A thread.
The fight between Humans and AI could be fun to watch.
I really like the idea to assign an identity to the chatbots, as they are trained and biased in a specific way, which is made more evident when they are encapsulated in a user profile.
Different chatbots can be trained in different ways, and have different characters, making them more interesting and nuanced for different situations.
All of this is great, and the user interfaces are changing very fast.
Quora is in grave danger, but I believe there is a glimpse of hope.
Poe’s user interface is much better than what I have seen from Microsoft or Google to date.
I believe the danger is real, but there might be a future for Quora and Stackoverflow - a coexistence of AI and human characters on the platform, fighting to provide the best answer possible — in the end the users will win, as the content quality bar will be set higher.