The Early Days of AI: A New Era Unfolding

The Early Days of AI: A New Era Unfolding

In the realm of artificial intelligence (AI), we often find ourselves viewing LLMs, Transformers, and diffusion models as mere continuations of past advancements. However, it is crucial to recognize that these developments mark an entirely new era and a significant discontinuity from the past. While previous waves of AI primarily benefited incumbents, the current landscape presents a unique opportunity for startups to thrive and create new market openings.

When discussing "AI" today, many business professionals perceive it as a continuum, encompassing the capabilities of the CNN/RNN/GAN world. In reality, AI represents a step function in terms of its novel capabilities and the products it enables. The launch of ChatGPT served as the starting point for mainstream recognition that AI is a game-changer. However, we must acknowledge that we are still far from reaching peak AI usage or peak AI hype, considering the lengthy planning and implementation cycles of large enterprises.

In these early days of AI, we can identify at least four waves that shape its trajectory. The first wave comprises companies native to the AI generation, such as ChatGPT, Midjourney, Character.AI, Stable Diffusion, Github copilot, and others. These early launches have gained significant revenue and user traction, establishing themselves as pioneers in the field.

The second wave, which we are currently witnessing, involves early adopters from startups and fast mid-market incumbents. These are the first startups to build their businesses on top of advanced AI models like GPT-3.5/4. Examples include Perplexity, Langchain, Harvey, and others. Simultaneously, a small number of founder-led multi-billion-dollar companies like Navan, Notion, Quora, Replit, and Zapier have rapidly launched AI-powered products and serve as the early adopters of this wave.

Looking ahead, we anticipate the third wave of AI startups to emerge soon. This wave will bring forth new formats such as voice and video, expanding the use of natural language in various verticals and novel ways. Additionally, we can expect advancements in infrastructure that support these startups' incremental experiences. Companies like Eleven Labs, LMNT, LFG Labs, Braintrust, and many more will contribute to the evolution of AI applications.

The fourth wave, projected to arrive around 2024 or 2025, will witness the first major enterprise adopters of AI. Due to the lengthy planning and build cycles inherent in larger companies, we can anticipate the launch of fully-fledged AI products, rather than mere prototypes or demos. This wave will extend beyond industry giants like Microsoft, Adobe, Google, and Meta, as other enterprises begin to leverage AI in their operations.

One remarkable example of AI's potential lies in Google's MedPaLM2 model, which outperforms human physicians to such an extent that involving medical experts in refining the model actually diminishes its performance. This finding highlights the immense power and capabilities AI possesses, even surpassing human expertise in specific domains.

Now, let's delve into some actionable advice for building AI-powered solutions. It is crucial to recognize that AI is a new tool that aids in solving customer problems previously deemed unsolvable. However, the fundamental dynamics of your business remain unchanged. AI technology alone does not create a competitive advantage or a moat. Instead, real power and durability stem from traditional sources such as economies of scale, network effects, counter-positioning, switching costs, and brand recognition.

It is essential to differentiate between growth and hype. While hype may garner attention, it is ultimately growth that determines long-term success. These two variables are independent, and focusing on genuine growth should be your priority.

When engaging in discussions about AI, it is common to hear experts debating alignment, defensibility, and AGI (Artificial General Intelligence). However, as builders, it is crucial to focus on the practical aspects of implementation. Many AI builders discuss where to find affordable GPUs, emphasizing that excellence in any field is primarily about hard work and attention to detail. The initial "big idea" serves as a seed, but how it unfolds and is executed ultimately determines its success.

In conclusion, the early days of AI represent a new era of technological advancements. Rather than viewing AI as a mere continuation of previous capabilities, we must recognize it as a distinct step function. As AI continues to evolve, we can expect waves of innovation, with startups, mid-market incumbents, and eventually larger enterprises embracing its transformative power. To thrive in this landscape, it is crucial to focus on genuine growth, leverage traditional sources of competitive advantage, and execute meticulously. The AI revolution is underway, and now is the time to seize the opportunities it presents.