post

Footer About

Footer Contribute

Footer Get Involved

Footer Links

Mind the Product

Career

Practical guide to move from IC to manager in product management

Artificial intelligence (AI)

How to build AI products that users trust: 3 core focus areas to stay user-centric

B2B product management

Artificial Intelligence (AI) & Machine Learning (ML)

Building Product Teams

Business Strategy

Data Driven Product Management

Digital Transformation

Diversity and Inclusion

Minimum Viable Product

OKRs

Prioritisation

Product Design

Product Research

Product Development Process

Product Ethics

Product Discovery

Product Leadership

Product Management Career

Product Strategy

Stakeholder Management

🌎 <a href="https://www.mindtheproduct.com/world-product-day/" style="color:#FFFFFF;text-decoration:underline;">Happy World Product Day!</a> Be sure to celebrate product, yourself, and your peers today! 🎉

Connecting fans to creators at Spotify (Dariusz Dziuk, Product Lead, Spotify)

Mistakes, learnings, and a product person’s relentless journey into leadership

Leading product ops at Delivery Hero with Koen Dubelaar

Everything wrong with the product management job market. And how to fix it

René is an experienced product leader who spent the last few years helping several SaaS start-ups in the creator space. He is a self-professed product geek, developer, blogger, and outdoor adventurer. He is passionate about creating high-quality products users love. His background includes roles in development, consulting, project management, and enterprise agility.

Rene Rosendahl

To state the obvious: “AI” is all the rage nowadays, and for good reason. We are probably experiencing the next major technological breakthrough in real-time, and this should be an exciting journey.
As Product Managers and <a href="https://www.mindtheproduct.com/whats-a-product-leader-anyway/" target="_blank" rel="noopener">leaders</a>, we obviously have a lot of questions. On the one hand, we worry about how this will affect our jobs. On the other hand, we are left (if not expected) to figure out how to make AI part of our products.
Before attempting to answer these questions, we should first get a few things straight, however.
When we say “AI”, we tend to refer to recent innovations around large language models (LLMs) such as OpenAI’s ChatGPT and Google Bard. This form of AI, which is also referred to as “generative AI” is made available by external providers who have spent years and lots of data to train these models. Those models run on their own (cloud) infrastructure and can be accessed by<a href="https://platform.openai.com/docs/api-reference"> </a><a href="https://platform.openai.com/docs/api-reference">APIs</a> these providers have made available. The output is largely text- (and recently also image and video) based. The term “AI” in general is a much larger set of technologies and approaches, not just generative AI.
Then there is “Machine Learning” (ML), which is technically a subset of AI. ML has been around for years and has already found many useful applications which have made inroads into our day-to-day lives, oftentimes without much fanfare (Amazon recommendations, anyone?). ML tends to be more focused in its application compared to generative AI and is great at dealing with numerical and quantitative data. Unlike the big LLMs, companies can create, train, and run their ML models on their own infrastructure with their own proprietary data. With that, there are much fewer external dependencies.
Getting back to how “AI” impacts Product Managers, there are really several parts to this answer:
<h1>1. A Tool to make life easier</h1>
This is the easy part: Like for many people, generative AI has the potential to make a product manager&#8217;s job easier and save time and effort. First off, there is content generation, which can help with anything from a release announcement, marketing copy, to internal documentation. Generative AIs can also be a valuable resource to aid with various research activities and supplement or even replace good-old Google. Then there is ideation to get creative juices flowing and provide ideas and starting points. Lastly, generative AI is making a foray into prototyping and testing of new ideas.
<h1>2. Integration into your product</h1>
In addition to using AI for many tasks, product managers will want to figure out how to <a href="https://www.mindtheproduct.com/a-deep-dive-into-how-ai-can-help-product-managers-succeed/" target="_blank" rel="noopener">integrate it into their products</a>. It stands to reason that sooner or later, most products will apply AI in some way, shape, or form and it may very well eventually become user expectation that a product be smart or intelligent. This task of integrating AI is more challenging as it requires a sense of how it works, an understanding of existing user problems it may address, and a degree of imagination and foresight as to how AI may evolve over time and how we may use it to create new capabilities.
These are still the early days of generative AI and most apps that already use it have not gone far beyond providing in-app prompts to help generate content. For many applications, it currently ends there and further interactions with the generated content are often not even possible. And therein lies the challenge: Generative AI is still largely text-centric and its output is unpredictable and unstructured, which makes it very difficult to create meaningful and interactive UIs around it. In a way, today’s mode of interaction is somewhat reminiscent of the early days of DOS with its text interface and command prompt. I believe one challenge for the future is how to evolve the underlying APIs to create more meaningful interactions that are not solely based on text.
One vector of evolution that’s already playing out is the emergence of<a href="https://platform.openai.com/docs/plugins/introduction"> </a><a href="https://platform.openai.com/docs/plugins/introduction">plug-ins</a> for LLMs, which allow 3rd parties to provide functionality to extend the natural scope of these generative AIs. Some early adopters have already started to create plug-ins for things like travel, shopping, and restaurant reservations. With those, ChatGPT can call out to these companies’ plug-ins to take actions beyond the generation of content, such as making restaurant reservations. Whether or not this architecture will be successful remains to be seen, but plug-ins are one way to integrate your product with ChatGPT (vs integrating an LLM into your product).
Lots of companies are currently exploring generative AI and many MVP and “me too” product features are starting to emerge, which is a good and logical start. More often than not, these take the shape of a text field with a prompt resulting in an output of generated text. I consider this the first baby steps of AI integration and productization. This is only the beginning and AI integration will have to mature quite a bit from here. In the end, it will likely come down to solving real customer problems via AI in a seamless way as opposed to just being able to claim “our product has AI features”. Apart from the initial AI appeal and excitement, adding real value will always outlive early hype.
Unlike generative AI, machine learning is a much more established path which might receive renewed interest due to the recent AI excitement. <a href="https://www.mindtheproduct.com/machine-learning-for-product-managers-a-quick-primer/" target="_blank" rel="noopener">The use cases ML</a> can facilitate are things like recommendations, classification, natural language processing (e.g. sentiment analysis), fraud &amp; risk assessment, and image recognition. While ML has been around longer and is also always improving, the rate of innovation here is certainly lower,  so it’s more approachable from a product and technology perspective (albeit much less sexy).
<h1>3. AI will take over the world and, with that, Product Management Jobs (just kidding)</h1>
Will AI impact many people and their jobs? Absolutely. Will it make PMs obsolete? Obviously, we don’t know, but I doubt it. Like many other groundbreaking innovations before, it will cause shifts and adaptation. But being a high-performing PM requires a broad set of skills (technical, product, customer, interpersonal, business-related, etc.), so I find it hard to believe that we’re anywhere close to simply automating and doing away with product managers. (Now, if you’re functioning more as a project manager-type product manager or a “Product Owner”/analyst gathering requirements, writing them up for the dev team, and reporting status back, I would be a little more concerned and look into evolving your skill set, regardless of AI.)
In other words, in a world of generative AI, the product job will &#8211; like many others &#8211; change and evolve. Evolving with and because of technology is part of being a product manager or really any type of knowledge worker, so enjoy the journey!
<h1>Takeaways</h1>
This is an exciting time and AI will certainly drive rapid change in technology, product management, and even society and the economy. As product managers, we’re not only users and beneficiaries of this new tech, we also have the responsibility of figuring out what it means for the products we’re in charge of.
As things move very quickly, we must remain curious and do what we can to learn. Part of that is using the new tools on a regular basis and to our benefit. We should also make the effort to understand the technical underpinnings &amp; principles &#8211; at least at a high level. We should think about how these new capabilities may benefit our products and how they could be integrated in an intelligent and seamless manner. (Hopefully, this means something more than just surfacing a ChatGPT prompt in our app.)
Lots of companies are trying to figure out how to use AI nowadays. Sooner or later users may experience “AI fatigue” since everyone will be touting “they have AI” in their product. Let’s remember that, at the end of the day, it’s not AI that wins, but creating lasting customer value and solving real problems &#8211; with AI (or without).

core/freeform

How Google makes AI work at Enterprise scale – Miku Jha (Director, AI/ML and Generative AI, Google Cloud)

LLM workflows for product managers: 3 key takeaways (Niloufar Salehi, Assistant Professor at UC Berkeley) – ProductTank SF

The future of product management: Insights from ProductTank San Francisco

Product lessons learned making early moves with AI in media: Lindsey Jayne (CPO, Financial Times)

A year with ChatGPT and product innovation: Navigating the AI landscape

How to keep your head about generative AI (when everyone is losing theirs) by Claire Woodcock

The AI effect: Transforming the world of product management

Rene Rosendahl

Read more

Comments