Google Product VP on AI Product Design (Part 3)

Looking at the evolution of AI product design, machine-based approaches are being combined with human-based approaches. The early algorithms set the stage for where we are now, but the introduction of human interactions like tagging took AI products to the next level. Ultimately, we arrived at a mid-point that balances technology and humanity to create products that connect with more people.

Today, we have access to more data than ever before. In addition, this data is driving the creation of sophisticated algorithms that are able to learn and adapt over time. For example, deep learning and the open web have opened the door to solving problems that used to be untouchable. Furthermore, the continuous acceleration of computing power increases our capacity to introduce breakthrough products.

While all of these advancements are exciting, it’s important to understand AI’s limitations. I will be the first to admit that Google Photos fails frequently. That said, we have a virtuous feedback loop that allows us to learn quickly and solve problems. Ultimately, people are still responsible for correcting these mistakes. However, these learnings allow our algorithms and systems to get smarter over time and increase their capabilities.

Our ability to continuously learn is driven by making data collection as effortless as possible. In other words, data needs to come from the real world rather than a controlled laboratory. For example, when you drive a Tesla in auto-steer mode, the car will capture every time the driver corrects their steering. Drivers aren’t touching the wheel because they care about making Tesla better. Instead, they are reacting to real-world driving conditions, which in turn makes the technology that much smarter through data collection.

Most importantly, the scope of AI products is reaching a wider audience than ever before. This technology isn’t just for big tech companies. For example, open source projects are turning high school students into full-scale developers. Simply put, the data we have combined with the human element is turning AI into a tool that anyone can use on a daily basis.

 

Click here for Part 1

Click here for Part 2

About the speaker
Brad Horowitz Member
Provide your rating for this post
If you liked this post, please use the buttons to the left to share it with a friend or post it on social media. Thank you!

Leave a Reply

Read more

AI Product Design: Data + Humanity

Human interaction has impacted AI product design by creating new data points to optimize products and make user interfaces more intuitive.

Juniper Networks Head of Product on AI & ML Product Design, Development, and Adoption

Juniper Networks Head of Product discusses AI & ML product design and how PMs can use the technology in product development.

AI & ML Product Design, Development, and Adoption (Part 2)

Juniper Networks Head of Product talks about the problem-solving path most PMs take and why AI and ML are not always the best solutions.

Sign-in / Register for Free

Don’t be left behind in your career. Join a growing community of over 500K Product professionals committed to building great products. Register for FREE today and get access to :

  • All eBooks
  • All Infographics
  • Product Award resources
  • Search for other members

Coming soon for members only: personalized content, engagement, and networking.