We are very proud to collaborate with Keiichi Matsuda and his team at Liquid City on this vision of the future where AI and AR come together to enrich and enliven the real world.
What I think is important here is the vision for embodied AIs that support our lives in the real-world in a transparent and collaborative way. There isn’t a singular, omniscient AI in this film, but rather several embodied AI agents assisting the humans at the center of the story, each agent with a role to play, transparent and open in offering help where appropriate, collaborating with other AIs and humans alike, and dismissable by our human protagonist when they aren’t needed.
This is a world where humans leverage AI to be better humans, where the information we share and the help we get remains firmly in our control. We think it’s a vision worth sharing and pursuing.
jh
By Keiichi Matsuda, Liquid City
In June we released Wol, an owl with an AI brain, who teaches you about its habitat, the redwood forests of California. Wol has now had over 70,000 conversations with people through Meta Quest Pro, and on iPhone and Android.
This project started as an experiment, building a character around a LLM-powered chatbot. We quickly saw its potential, and built Wol as a proof-of-concept fulfilling a real use case - personalised education.
We recently released our new short film ‘Agents’, that suggests a wider vision for how agents like Wol could become part of our everyday life.
In Agents, there are agents for every conceivable purpose. Nav gives you recommendations for places to go, and leads you there. Vibes is your personalised music and IoT lighting selector, responding to your mood. The well-being listens to your long term goals, and helps you stick to them. There are agents that process payments, agents that look after your plants, personal trainer agents, recipe inspiration agents.
It’s a different vision for computing - there are no screens, no buttons, no floating panels (well maybe one or two). Apps are reimagined as animated characters that you can talk to. Maya, the film’s protagonist, interacts with technology in the same way as she might interact with friends or colleagues.
Two weeks after its launch, people have been talking about it as a positive vision. It’s definitely more optimistic than its spiritual predecessors, HYPER-REALITY and Merger. But what makes it feel more desirable? Here are some of my thoughts.
Connection to people and place
AR is computing that happens in the world around us - not a separate dimension of personal screens, but a shared world where virtual beings coexist with us. Agents can be owned by you, but you can see other people’s agents too. There might also be agents who ‘haunt’ particular places, or are tied to specific objects. Maya’s interactions with technology connect her to the people and places around her.
Focus on ecosystem
Maya has a pantheon of agents. Some are made by companies, some are simple agents that can be made by a single developer. Maya is not a developer but can make agents herself. This is not a view of the future where monopolistic super-intelligent AI is taking over - it’s a more humane vision where AI and humans can peacefully exist, and people are producers as well as consumers.
Trust is prioritised
Managing trust is not handled by today’s interfaces. Click ‘accept’ or ‘decline’, or go into menus that are difficult to understand and easy to ignore. In the world of Agents, we can rethink this. When Maya is spammed by the ‘local guide’ agent, she’s able to see that its her trusted agent Nav leaking her data. She can sever their connection by simply pulling them apart, thereby adjusting her privacy settings. Agents are not intrinsically trustworthy - they all have different motivations, just like people or companies in the real world. You curate your inner circle of agents according to your needs, as well as your taste, and how much you trust the makers of the agents.
A desirable future isn’t one where every problem is solved, or that everyone’s interests are perfectly aligned. It’s one where we have the agency to choose, to exercise control in how we live our lives.
It feels like the whole world is talking about advances in AI, and that’s sparking fear - about job replacement, killer robots, losing control. These are real concerns, but we should also be talking about our hopes for AI. What kind of world would we like to see? How would we like to see AI integrated into our daily lives?
Agents is a speculative film - the technology isn’t quite at the level shown in the film yet. Talk to Wol and you may get stuck in a polite loop of pausing and interrupting, like a long distance video call. But almost everything can be built right now - we are not waiting on any science fiction technological breakthrough - we have the tools today to make this vision into a reality.