To Investors and Entrepreneurs: My Predictions
Jan 2025
Over the past six months, these insights have helped me navigate key decisions in my projects. As we kick off the new year, I thought it’d be worth sharing a few predictions that might guide investors and entrepreneurs in the months and years ahead.
Prediction 1: Startups without network effects are doomed
Most startups focus on solving a specific problem by writing reusable logic to sell. But with AI agents, that model is on the way out. AI will handle business logic far better than any startup ever could (OpenAI’s o3 is already hinting at how these machines will outperform humans in almost any task).
The only thing AIs can’t compete with is network effects: the more users, the better the product gets.
That said, in the short term, startups leveraging third-party LLMs for niche, well-defined use cases can scale fast, attract customers, and rake in profits at an insane pace. For investors chasing quick exits and entrepreneurs after rapid gains, this is still a viable play—for now.
Prediction 2: AI-powered scientific discovery is the exception to Prediction 1
We’re about to see the rise of labs on the edge of scientific breakthroughs, creating massive value for society. AI will be fundamental to every major discovery moving forward. If the recent Nobel prizes in Physics and Chemistry are any indication, this is just the beginning.
Prediction 3: Hardware will define the next competitive edge
While software eats the world, hardware is quietly emerging as a critical differentiator. As AI systems become increasingly computationally demanding, there’s a growing need for innovative hardware solutions tailored to AI workloads. Companies like NVIDIA have already shown how hardware advancements can reshape entire industries, and we’re just scratching the surface.
Startups working on custom silicon, edge AI devices, and quantum computing hardware will have a unique opportunity to create defensible IP and dominate their niches. Unlike software, hardware takes time and capital, but the barriers to entry provide long-term competitive advantages.
For entrepreneurs, focusing on specialized hardware for specific AI use cases could be a game-changer. For investors, this is an area where patience could pay off big, as the demand for more efficient, scalable, and domain-specific hardware continues to rise.
Prediction 4: AI safety will become a central investment focus
As AI systems grow more capable, ensuring their safety will become a critical challenge. Governments, organizations, and consumers will demand robust mechanisms to prevent misuse and unintended consequences. Startups specializing in AI safety tools, interpretability frameworks, and fail-safe mechanisms will attract significant funding.
This is not just a moral imperative—it’s a market opportunity. Investors who back ventures working on scalable AI safety solutions will not only contribute to a safer future but also position themselves at the forefront of an industry set to grow exponentially.
Entrepreneurs should consider how they can integrate safety and reliability into their AI systems from day one. This will not only differentiate their products but also future-proof their business against potential regulations and societal pushback.
Prediction 5: Crypto will keep growing
With geopolitical instability on the rise, Bitcoin’s reputation as independent and trustworthy will continue to attract both governments and individuals. It remains the safest and most reliable store of value.
And if quantum computers ever become a threat, the community will adapt with a workaround—just like it always has.
For entrepreneurs, there’s untapped potential waiting to be explored. For investors? Just hold onto some BTC—it’s a long game. (This is not financial advice.)