The Future of AI: Predictions and Challenges

The Future of AI: Predictions and Challenges

Artificial intelligence (AI) is rapidly advancing, and it is seen as the foundational platform from which numerous advancements will be made across all industries. However, before creating AI, it is important to know what we want answered and to reframe the problem statements. Building a list of questions and giving the whole list a structure can help create good user interview questions.

From massive leaps forward in large language models and multimodal models that move between image and language, to applications that significantly extend the capabilities of scientists, AI is set to change the world. There will be new business opportunities based on the APIs that very large models will create. The challenge is to create distinctive businesses on them and to think about how to create an enduring differentiated business.

There will be a small handful of fundamental large models out there that other people build on. However, there will be a middle layer that becomes really important where companies will create a lot of enduring value because they will have a special version of it. The alignment problem is how to build AGI that does what is in the best interest of humanity. The quality of language models we’ll see in the coming years will provide a serious challenge to Google for the first time for a search product.

The marginal cost of intelligence and the marginal cost of energy are going to trend rapidly towards zero, surprisingly far. The metaverse turns out to be more like something on the order of the iPhone, a new container for software and a new computer interaction thing and AI turns out to be something on the order of a legitimate technological revolution. The economic impacts are huge, and figuring out the answer to those questions is going to be huge.

Eventually, AI will be able to do the whole creative job. Startups will train their own models, just not from scratch and will take base models that are hugely trained with a gigantic amount of compute and data, and then they will train on top of those to create the model for each vertical. Those startups will be hugely successful and very differentiated startups there. However, the world also just doesn’t make enough chips.

In conclusion, AI has the potential to change the world as we know it. It is important to create good user interview questions, reframe problem statements, build a list of questions, and give the whole list a structure. There will be new business opportunities based on the APIs that very large models will create. The challenge is to create distinctive businesses on them. The alignment problem is how to build AGI that does what is in the best interest of humanity. The economic impacts are huge, and figuring out the answer to those questions is going to be huge. Eventually, AI will be able to do the whole creative job.