Character AI's Journey: From Not Needed to Highly Valued

Challenging Conventions: How Character AI is Carving Its Niche Amidst Tech Titans (and Can It?)

Introduction

It's a classic underdog story: Character AI, despite its innovative edge, which went underappreciated during its founders' tenure under Google, now faces an uphill battle against larger, well-funded rivals like Meta. They exemplify a missed opportunity for one giant – Google, now seemingly eager to rejoin the fray; and serve as an inspiration for another – Meta, possibly leading to Character AI's eventual crash, given how hard it is to compete with a company of over 3 billion users, especially having almost no revenue, nor a sustainable business model.

What put it on our list? Only eight months ago, in March, the startup raised $150 million in a funding round led by Andreessen Horowitz (a16z), achieving a valuation of over $1 billion. Recently, there's been news of Google's interest in investing 'millions', potentially elevating the company's valuation to over $5 billion. This situation turns Character AI into a potentially significant player in our AI Unicorn family list.

However, there's also a profound risk of failure.

So what is Character AI? Let’s take a closer look at what we already know about them:

1. Google didn't pay enough attention – the starting point of Character AI

2. De Freitas' childhood dream to talk to computers: the vision behind Character AI's mission

3. Character AI's financial situation: from seed rounds to Google's multi-million dollar interest

4. Character AI’s main offer: tapping into the Artificial General Intelligence (AGI) concept

5. Tech behind Character AI’s product

6. How does Character AI make money?

7. Conclusion

Google didn't pay enough attention – the starting point of Character AI

During their time at Google, Noam Shazeer and Daniel De Freitas were instrumental in the projects that could have emerged as significant as ChatGPT did. In 2017, Noam Shazeer co-authored the seminal paper 'Attention Is All You Need,' laying out the groundbreaking Transformer architecture, which catalyzed the generative models uprise. He was also the first to scale Transformers to supercomputers and pioneered large-scale pretraining. Drawing on this attention paper, Daniel De Freitas, in 2020, initiated and led the Meena project, which was later renamed LaMDA. This 2.6 billion parameter, end-to-end trained neural conversational model, unveiled in 2021, was capable of… chatting about anything.

Sounds familiar, doesn't it? However, Google held back on these innovations. Shazeer’s proposal for a chatbot never materialized at Google, hindered by concerns over potential missteps in its responses. Google's executives repeatedly dismissed the idea, at least once stating that the AI didn’t align with the company’s standards for safety and fairness in AI systems.

Consequently, at the end of 2021, both Shazeer and De Freitas resigned from Google and founded Character AI – a startup focused on creating chatbots capable of mimicking famous personalities.