The Evolution of Machine Learning - Dan Jeffries | Productive AI Podcast

The Evolution of Machine Learning - Dan Jeffries | Productive AI Podcast

In this episode, I speak with Daniel Jeffries. Daniel is a science-fiction author, engineer, futurist, thinker, blogger, systems architect, speaker, crypto nerd, AI evangelist, world traveler, beard-master, and overall renaissance man.

Today we talk about a variety of topics including: why billionaires going to space is good; why how to make better predictions; how COVID will have long-term positive consequences for society; where we are in the long arc of AI; how the model development lifecycle supports and does not replace the software development lifecycle; where we are in terms of understanding MLOps; choosing between end-to-end and best-of-breed ML tools and platforms; what the AI Infrastructure Alliance is and how it’s helping shape the future of ML Platforms; and what to think about when deploying AI/ML in your organization. It’s a long and great conversation. Enjoy the ride!

Get in touch with Dan:

Amazon Author Page:
Patreon Page:

Listen to the full episode on Simplecast

Watch the full episode on YouTube

00:00 Introduction
02:15 Why billionaires going to space is a good thing
04:13 Dan’s thoughts on the Foundation series
05:55 Predictions - good and bad - that Dan has made
10:19 Thoughts on Kai-Fu Lee’s “2041”
12:40 COVID’s long-term impacts on our society
21:06 Where are we now in the arc of AI?
27:12 This is still the early adopter phase
29:42 Is AI really eating all software?
31:36 The model development lifecycle vs. the software development lifecycle
33:07 MLOps is still evolving as a term and as a practice
36:07 MLOps is not just DevOps brought forward
39:15 ML Platforms: End to end or best of breed components? (Or a blend?)
40:34 The only end to end solution that exists is in the minds of marketers
44:38 There is no LAMP stack for machine learning...yet
47:54 What is the AI Infrastructure Alliance
57:13 Blueprints and design patterns - making sense of the ML platform and tools space
1:05:35 Platform rationalization and maturation is coming but it’s not here yet
1:07:30 How does a customer buy from members of the AIIA?
1:11:45 Education is critical to long-term success
1:17:15 As always, finding the right tool for the job is important
1:21:45 There are two kinds of machine learning: basic and revolutionary
1:24:35 Wrap-up

Watch just the part you want on YouTube

Watch some of the best quick clips!

AI is in the early adopter stage:
MLOps is still evolving:
There are no end-to-end ML platforms yet:
There is no LAMP stack for ML yet:
AI Infrastructure Alliance goals:
ML Platform blueprints:
Two kinds of machine learning: