You recently spent some time at Ontario Teachers' offices in Silicon Valley. What are some of the trends you're noticing?
ALD: I was at Stanford University in the late 1990s and early 2000s, right when the first Internet boom was happening. Silicon Valley was an extraordinary place to be, with no limit to your imagination. Today, it’s the first time since then that I feel that same energy, with what's happening in AI, this explosion of data, the tools around it and the cooperative ecosystem. There's still a lot of research that needs to go into AI to create more breakthroughs, but it is already incredibly exciting.
Thanks to the wide adoption of cloud and the ever increasing compute power and mobile technology, we are seeing a continuous explosion of data. Businesses now universally understand how valuable data is. Data can give you insights on your customers, your employees, your productivity and your competitors. It can help you improve your user journeys, your conversions and your sales pipeline. But to do so you need to first collect it, observe it, access it and analyze it. The infrastructure for our data and the tools that are required for businesses to extract value from it is a huge area of innovation.
Data is also the lifeblood of AI. For AI to work well — to give you high quality output — you need high quality data as an input and you need lots of it. It is no coincidence that we are seeing a step change in the AI world as a result of the accessibility of data and compute power. Generative AI is one form of AI which has enabled us to use unstructured data at scale as inputs for lines of text, articles and tweets, thus increasing the models’ understanding of the world. It has also enabled us to converse with machines to question, understand and interact with them. While they still have many weaknesses and limitations these will be addressed with time and with it, we are witnessing the creation of a new ecosystem of companies.
When the iPhone first came out, we were like, "That's great, but what does it really do for me?" But a couple of years later, a whole ecosystem of applications was built for the iPhone; millions of them, because the barrier to coding for the iPhone relative to the PC was much lower. In comparison, today, large language models (LLMs) give you the ability to use language to access certain logic and coding thus enabling millions of coders and innovators. It's still imperfect, but it's a huge step forward.