The audio and video are in Spanish. These notes are a translation.
ChatGPT has been the star of news covers and social media for the last 2 months, but artificial intelligence is a technology that has been in development for several years.
Why has OpenAI suddenly stolen the spotlight, and what opportunities do the latest advances in artificial intelligence bring for investors, founders, and corporations?
Today I spoke with Miguel Paredes, Vice President of Data Analytics and Artificial Intelligence at Albertsons, the second-largest supermarket chain in the United States. Before that, Miguel earned a PhD in Data Science and Artificial Intelligence from MIT.
Miguel is also an angel investor and a Venture Fellow at AI Fund, an artificial intelligence venture capital fund founded by Andrew Ng, a Stanford professor, cofounder of Coursera, and founder of Google’s AI arm.
We talked about AI models, what is happening behind the popular ChatGPT, and what opportunities Miguel sees with this technology.
Youtube Timestamps*
[2:17] - History of Artificial Intelligence
-
How did Miguel arrive at the fascinating world of startups and artificial intelligence?
-
What have been the most significant advances and the companies that have brought us to the modern era of artificial intelligence?
-
There are many advances, such as language models, neural networks, transformers, and reinforcement learning, that go back more than 5 years. They are not recent developments, but he understands that they play an important role as a precedent for OpenAI’s current work. Miguel briefly explains what they are (in simple terms) and why they are so important.
-
OpenAI has stolen the spotlight. What is that special thing it has achieved? Is it the enormous amount of data used, the design of the models, the launch of ChatGPT with a friendly interface, or what?
[20:06] - The Technology Behind Artificial Intelligence
-
What actually happens during the training of an artificial intelligence model?
-
What happens inside these models when you ask them for something? What is the operation, and what is going on behind the scenes?
-
We have seen DALL-E 1 and 2. We have seen GPT 1, 2, 3, and soon 4, which, from what people say, is hard to believe. What happens between the model’s evolutions? What are the teams behind them doing to improve these things? Is it human work, or the model improving itself?
[34:06] - Opportunities in AI
-
In Silicon Valley, people talk about the idea that the opportunities for startups lie in applying artificial intelligence to specific sectors and problems, and not in the technology itself, because this technology will become an infrastructure service like the cloud, which will be dominated by the big technology companies such as Microsoft, Google, and Amazon. Does Miguel agree with this vision?
-
Let’s say Miguel is forced to quit his job and start a startup taking advantage of artificial intelligence: How would he approach the challenge? What things would he focus on? What would be a good mental model for creating a valuable company based on artificial intelligence?
-
What does Miguel recommend to founders who are in growth stages and already have data they could use with artificial intelligence?
-
Beyond ChatGPT. On a horizon of 12 to 24 months. What is Miguel following most closely? Where does he have his eyes, so that if something comes out, you are the first to know?
**[**53:03] - Rapid-Fire Questions
-
If he had to enter the world of artificial intelligence from scratch, what is the main piece of advice Miguel would give himself?
-
What has been the most influential book on Miguel’s path through the world of startups?
-
Changes Miguel would like to see in Latin America’s startup ecosystem
Episode Mentions
You can find Miguel on LinkedIn
Recommended book: Secrets of Sand Hill Road by Scott Kupor
Discover More from the Startupeable Podcast
Listen to every episode at startupeable.com/podcast.
Subscribe to the newsletter to hear when a new one comes out.
Follow Enzo Cavalie on X: @enzocavalie.