We set Quora’s original mission, to share and grow the world’s knowledge, in 2011, a time when software could not do much more reliably than move information around between people.

The first version of Quora did not use machine learning at all, instead relying on users to tag content, tag their interests, and manually filter for questions they could answer.

Over the years, as machine learning became more powerful, we used it automate this manual work. As large language models emerged, we started using them to generate questions, facilitating human answer writing.

In early 2022, we experimented with applying GPT-3 to answering questions. These experiments revealed that LLMs could be very useful in private, whereas Quora content was all public.

We also realized that people could often collaborate most effectively with LLMs through a chat interface. So, we created Poe, a product for people to chat with all kinds of LLMs in private.

Today Poe lets millions of people access knowledge and intelligence from all kinds of AI, including text, reasoning, image, video, and audio models.

People use it for very diverse tasks: they ask questions, they get help with their writing, they generate multimedia, and they create entire applications.

While some people are looking for knowledge, as the models get smarter and smarter, people want their advice, their collaboration, and more generally, their intelligence.

Quora’s public format and Poe’s private focus make them complementary in serving a wide spectrum of needs, but they function similarly: they both let people access knowledge and intelligence that helps them understand and affect the world.

They both aggregate knowledge and intelligence from many sources. They both focus on sharing what exists, rather than creating knowledge like a researcher would or creating AI like an AI lab would.

They are platforms, and as we build them we inherently focus on the most leveraged aspect of creating public goods: removing friction and unlocking the potential in what everyone else has already learned, created, or trained.

By sharing all this knowledge and intelligence, we help individuals, but we also hope to help society as a whole.

The more we enable collaborative thinking between humans, and between humans and AI, the faster ideas will diffuse, the faster technology will advance, and the faster the economy will grow.

Our ultimate goal is at this higher level: the ability for humanity as a whole, along with all artificial intelligence, to share knowledge, create technology, and work together to solve problems. We describe this as “collective intelligence”.

This is our new, expanded mission: to grow the world’s collective intelligence. Full blog post: https://quorablog.quora.com/Ex...

@adamdangelo As someone who has contributed a lot to Quora over the years, have to say I honestly have no idea what this thread means

@adamdangelo Congrats Adam! Love the new mission.

@adamdangelo Terrific, gels

@adamdangelo @poe_platform This is awesome

@adamdangelo Go on @Quorators Adam