What We Can Learn About AI from Jack Dorsey

At the recent JP Morgan Tech conference, Block co-founder and CEO Jack Dorsey shared his thoughts on the rapidly evolving field of Artificial Intelligence (AI). His remarks offer valuable insights into the current AI environment, including the value of open-source models, the importance of data, and the practical application of AI in solving real-world problems. Coming from one of the most experienced Silicon Valley CEOs and entrepreneurs, his comments also offer valuable insights for tech investors looking to capitalise on the AI gold rush.

Open Source: A Level Playing Field

"I'm so thrilled that there exists an open-source model competition for the space because it means companies like ours, whose first priority is not building AI models, can actually compete on a level playing field”.

Dorsey expressed enthusiasm about the rise of open-source AI. Specifically, he highlighted recent advancements upon the leaked Facebook LLaMA models to quickly approach the quality of OpenAI’s GPT-4.

"When I saw that leak from Facebook and how quickly the community iterated on it, getting to the point of being passed GPT-3 and approaching GPT-4 and the fact that it's completely open and we can self-host it, I'm like, damn, this is phenomenal."

This development, Dorsey argues, allows companies to tap into the knowledge and tools built by the AI community, bypassing the need to develop their models from scratch. For companies like Block, this presents an opportunity to leverage advances in AI without investing significant capital in high-risk R&D. 

The Power of Data

Dorsey emphasised the crucial role data plays in making AI models effective. "The true heart of what makes any AI model great is that data," Dorsey said. AI thrives on data, and companies like Block, with its specific data about both sellers and consumer financial activities, are well-positioned to harness its power.

This is something that has been lost in the current discourse. While foundational language models are awe-inspiring in their ability to perform a wide range of tasks, they can become even more powerful when fine-tuned on domain-specific data. From this perspective, companies with high-quality proprietary data, like Block, have a clear advantage.

From Hype to Practical Use

With the recent hype surrounding AI, Dorsey warned against companies rushing to incorporate AI for AI’s sake.

"I think there's a ton of hype right now, and I think there's a lot of companies being started that are going to fail because of that hype,"

The technology industry, he noted, often succumbs to the allure of trends, jumping from one buzzword to the next without fully appreciating their practical application or longevity. If anyone understands this, it’s Dorsey, who has been a long-standing proponent of Bitcoin and decentralised tech and has thus experienced the vicissitudes of the tech hype cycle better than most.

Instead, Dorsey urged companies to approach AI from a 'use-case' perspective, identifying specific problems they're trying to solve and then exploring how AI can be used creatively to address those issues.

In the context of Block, AI has vast potential to improve accessibility, increase efficiencies, and help the company operate more like an internet company.

A Bright Future

While Dorsey struck a cautious tone on the hype, his reflections on AI signal a bright future for the technology, particularly in finance. Open-source models are creating a new paradigm, allowing companies to tap into AI’s power without becoming specialists. Meanwhile, the effective use of data and a focus on practical applications promise to make AI a transformative force in business and beyond. It's a testament to the democratising power of AI and a glimpse into a more connected, efficient, and equitable future.