The future of AI might look a lot like Twitter

Michael Salmon shares his journey in creating Social AI, an innovative app where users interact with AI bots rather than real people. This idea stemmed from his extensive experience at top tech companies and previous attempts that didn't succeed due to limitations in technology. He aimed to mimic a real social network experience without the downsides, seeing his app more as an interface than a conventional network. Salmon compared his early attempts to using a nicotine patch for quitting smoking, trying to replace a harmful habit with a less harmful experience. With advancements in AI language models, he believes he can now offer users a richer experience that does not remind them of technology, striving for something that truly feels social and engaging.

In recent years, social media has changed drastically due to advancements in technology. Instead of using the internet to connect with others, people are now communicating directly with AI models, which can create an illusion of connection without real human interaction. This shift can be both beneficial and harmful, as it raises questions about authenticity and the role of human connection. The speaker reflects on their journey to develop a product that uses AI to provide an alternative to harmful social media, offering a space for people to express their thoughts instead of interacting negatively with the online world. They emphasize the importance of genuine human communication while acknowledging the growing need for support, especially in a time when many people feel isolated, particularly after COVID-19.

AI tools can offer valuable support in our lives, but they should be used to complement, not replace, human connection. Many people turn to social media and AI for advice in stressful situations without understanding how algorithms work. This can lead to confusion. A new approach is to combine AI's capability to provide multiple perspectives in a social-like format. Instead of just receiving one answer from traditional chat interfaces, users can experience a more natural interaction where they can post their situation and receive a variety of responses. For instance, during a travel dilemma, users can get different suggestions on how to navigate a tight schedule, fostering an engaging exchange that feels personal. This way, AI becomes a practical tool in decision-making while preserving the essential human element in communication.

Social AI offers a unique experience where users interact with bots that reflect various personalities, creating a dynamic environment for discussion. Unlike traditional social media, where users often feel pressured to present curated thoughts, Social AI allows for a more liberating experience. Users can seek different perspectives and even challenge their own views without the fear of judgment. Interestingly, while many reacted negatively to the concept of an app primarily made of bots, some found liberation in engaging with them. Many users chose contrarian bots, indicating a desire for honest debate and diverse viewpoints. This seeks to break down echo chambers, as they can confront ideas they might shy away from in typical social media settings. This newfound space may encourage people to explore and grow their thoughts and beliefs more openly.

Social AI is evolving as a fresh interface for interacting with AI systems, aiming to facilitate idea generation without the usual stakes of human interaction. It allows users to connect with AI personas that reflect their interests and preferences, creating a personalized feed. Users can engage with a variety of responses, prompting brainstorming and discussions based on trending topics tailored to individual preferences. This dynamic shifts traditional social media experiences by focusing on user-centric content, where AI curates information relevant to each individual. The goal is to blend creativity and exploration, resulting in an environment that feels safe and stimulating, encouraging deeper engagement with ideas through an endless feedback loop.

Interactions with social AI, like asking for movie recommendations, showcase both the strengths and weaknesses of AI systems. While some platforms provide a variety of responses, which can be helpful, others might disappoint with a single answer that doesn't resonate. It's essential to recognize that even in their early stages, these tools aim to enhance decision-making and engagement by offering multiple viewpoints. This reflects a more interactive and less rigid approach compared to traditional AI models, which typically give one response. As social AI evolves, it seeks to improve user experiences by accepting that not every answer will be perfect. Ultimately, users now navigate a landscape where they can choose from different opinions and recommendations, making the experience more user-friendly and dynamic. This adaptation process mirrors the growing understanding of social media algorithms, where users often brush off less relevant suggestions in favor of more engaging content.

People still use Google because it provides various links and perspectives, despite issues like paywalls that hinder access to information. There's a vision for a platform that allows users to explore multiple AI language models, akin to using Google but for AI answers. This platform would help users discover and filter their preferences for conversations with different AIs, ultimately leading to more meaningful interactions. Instead of simply chatting with one AI, users could navigate through various communication forms—like audio calls or chats—until they find the most relevant AI for their needs. The goal is to enhance the overall experience of communicating with AI, moving away from the limited interactions we currently have. The aim is for future technology to intuitively understand user preferences without requiring extensive setup each time.

Building a platform that allows people to think before expressing their thoughts is valuable. Many feel overwhelmed by the pressure to share on social media. A safer space for reflection and conversation can help connect those who may not have anyone to talk to. This platform can be beneficial for everyone, also by reducing careless posts on public platforms like Twitter. It’s heartening to know that even unexpected users, like the speaker’s mother, can find a way to use this technology to express their feelings. This shows that the design of such tools can reach people in unique ways, making a positive impact on how we communicate our ideas.