AI / ML is Eating Product Quickly

Hello Product Folks!

AI has been part of our daily routines for years…It just hasn’t made the headlines (everyday!) until now.

From streaming services like Netflix to virtual assistants such as Assisto, its presence is ubiquitous or even with autonomous vehicles with Waymo, where AI-driven computer vision enables safer navigation.

Some startups such as BenevolentAI use AI to analyze biomedical data to identify potential drug targets.

Others such as Envision Pharma apply AI on healthcare and clinical trial data to help optimize drug development processes and improve patient outcomes.

Okay…Enough of examples, now what’s in there for product managers…

1- AI vs machine learning vs deep learning

• First, let's start with the basics and clarify the distinctions between AI, machine learning (ML), and deep learning (DL). In short,
  
  - AI, or artificial intelligence, simulates human intelligence. It utilizes existing data to enhance predictive capabilities.
  
  - ML is a subset of AI focused on automating learning and improving experiences.
  
  - DL, on the other hand, extends ML by employing intricate algorithms and training models for more complex tasks.

a- What’s Generative AI (GenAI)

• We cannot talk about AI without GenAI these days, so let’s set the context.
  
  Gartner categorizes AI into data-centric, model-centric, application-centric, and human-centric AI. At this stage, we're at the peak of inflated expectations, focusing on generative AI and responsible AI.

  
  

  

  gartner.com

  
  
  Generative AI, led by OpenAI's ChatGPT, utilizes vast datasets to generate predictions and contextual responses.
  
  ChatGPT's success, reaching 1 million users in five days! (count me in there), democratizes the power of GenAI to mass consumers.

  
  Soon, AI will be deeply integrated into the product management practice, and we will no longer talk about it. Not the best analogy, but it almost felt like Cloud Computing in the early 2000s when it was cool to talk about it (aka buzz) - today, most products are Cloud-based.

2- What’s an AI PM in practice?

• Switching gears, let’s recap what a PM does. A quick definition for those unfamiliar with it.
  
  PMs collaborate with various teams like UX designers, engineers, sales, marketing, legal, and customer support to align information and garner support to drive their product strategy.
  

  Now, when dealing with machine learning (ML) product development, the complexity increases. PMs navigate through vast amounts of data, collaborate with new stakeholders ML and data engineers, and focus on new metrics like accuracy, precision, and recall.
  
  Unlike traditional product development, AI / ML products emphasize performance metrics over adoption frameworks.
  
  Development involves working closely with data scientists and ML engineers, analyzing patterns, and overcoming challenges associated with data quality and scalability.
  
  Estimating development timelines in this space can be particularly challenging due to the unpredictable nature of data-driven solutions.