The Africa Deep Tech Community recently hosted a thought-provoking session with Zak El Fassi, founder and CEO of Zaigood Labs, an experimental AI venture studio based in Oakland, California. Zak, known for pushing the boundaries of conventional thinking in AI, led an engaging discussion that explored the untapped potentials of AI through a post-GPT4 lens.
AI in a Post-GPT (Generative Pretraining Transformer) 4 World
Zak began the conversation by illustrating his view of the current state of AI and its future trajectory. The GPT-4 model, the latest iteration by OpenAI, is radically different from its predecessors. He explained how the switch from GPT-3.5 to GPT-4 was akin to the shift from magic tricks to a true representation of intelligence.
GPT-4’s larger context window allows the model to compute much more sophisticated data, enabling it to deliver valuable insights and creative solutions. However, the expanding context window – the metaphorical ‘brain’ where AI holds a specific amount of data during a session – also presents challenges such as higher costs, slower outputs, and more complexities.
The Potential of Large Language Models
Exploring auto GPT and GPT engineer, El Fassi showcased how existing applications of AI, like large language models (LLMs), are revolutionizing different domains. From generating complex codes to simulating in-depth project management strategies, these AI tools can help us increase efficiency, spark creativity, and rethink how we approach problem-solving.
Zak also highlighted that beyond simply automating tasks, LLMs empower individuals by giving them superpowers. These tools, he argued, are enablers rather than replacements for human capabilities.
AI and the Evolution of Media
Another significant area where AI is making strides is media. Zak divided digital media into three broad categories: text, audio, and images. He explained how advances in large language models have progressed from text to more complex mediums like audio and images.
He illustrated this evolution with the use of the newly-launched Dall-E image generation API by OpenAI. By leveraging AI, a new era of creative expression was opened up – from the generation of imaginative text to the creation of strikingly detailed and contextually accurate visual imagery.
The Future of AI: Beyond Tools
During the session, Zak challenged the conventional AI narrative that focuses on AI as a tool for efficiency. He led the audience to ponder on a critical question: What happens when AI technology becomes an integral part of our lives, and even our identities?
Looking ahead, Zak posited the idea of language models serving as a personal ‘AI avatar’ acting on its users’ behalf, raising fascinating and somewhat unsettling questions about the future of our digital identities.
By exploring these ideas, Zak prompted us to consider how AI might reshape not just our relationship with technology, but also our perceptions of what it means to be human in an increasingly automated world.
Conclusion
Though still in the experimental phases, the potential of AI is awe-inspiring, and sometimes, overwhelming. As Zak El Fassi demonstrated, we’re only just beginning to unlock the full capabilities of AI. As we navigate through these uncharted territories, sessions like these are vital in fostering meaningful conversations and harnessing the potential of AI to drive significant change.