Scrum.org continues to research the impact of AI on agile product development. As a part of this work we have begun creating a series of podcast episodes with the latest being: ‘Beyond the Code: Vibe Coding, AI Agents and Scaling Autonomy with Tomasz Maj of Odevo’. There are many things that we are learning during this research that are being introduced into our materials. I will share more of this learning in future blogs, too.
In this blog, I discuss a phrase that has come up in these interviews and in general when talking about AI.
"AI is the next printing press"
This phrase describes the impact of AI on society and how, ultimately, it will change the world in unimaginable ways, like the printing press.
The Impact of the Printing Press
Apparently, the introduction of the printing press was a collection of innovations, but most people focus on the impact of Gutenberg's movable type in 1450 as the disruptive change. This is similar to AI. AI, as we see it today, is actually the continued evolution of digital technology since the microprocessor was created in 1971.
Before Gutenberg’s innovation, books and written knowledge were a luxury good. Because production was limited to scribes, only the elites, religious institutions, and wealthy patrons could afford it. The delivery process was cumbersome and error-prone, and it took a long time for a new product to ship.
The Gutenberg printing press changed this:
- Mass production at low marginal cost: One setup could quickly and cheaply yield hundreds or thousands of identical copies. This turned knowledge from a bespoke, scarce product into a standardized, abundant one.
- Faster and wider distribution: Books, pamphlets, and early newspapers spread via emerging trade networks (e.g., Venetian shipping routes). Ideas reached distant markets rapidly, creating the first mass-communication channels.
- Democratization and new markets: Literacy rose over time as access broadened. New industries emerged—publishing, bookselling, papermaking—while old roles (scribes) declined. It enabled "product variants" like affordable editions or localized translations.
- Downsides and adaptations: Initial fears of job loss, errors in copies, and misinformation (e.g., pamphlets fueling religious and political upheaval). Society adapted with new skills (proofreading, editing), regulations, and literacy education. In the long term, it accelerated the Renaissance, Reformation, and Scientific Revolution by enabling rapid iteration of ideas.
AI is the new ‘cognitive’ printing press
Access to AI and automation offered by platforms such as Claude, Gemini, and ChatGPT enables knowledge workers to rapidly create and adapt products by democratizing capabilities. I am not a front-end developer, but with AI, I can build a fantastic user experience.
- From bespoke/scarce to abundant and customizable: Just as the printing press made books affordable beyond elites, AI lowers barriers to product creation. A small team or founder can now generate designs, code, marketing copy, or even prototypes far faster than traditional R&D cycles. Tools like generative design AI or code assistants compress months of work into days. For physical goods, AI optimizes CAD models or supply chain forecasts, enabling on-demand manufacturing (e.g., via 3D printing integration).
- Speed and efficiency in the delivery pipeline:
- Design and ideation: AI accelerates iteration—simulating thousands of variants, predicting performance, or personalizing features based on data. This is like movable type, allowing quick changes to typesetting versus rewriting entire manuscripts.
- Production and operations: AI drives predictive maintenance, workflow optimization, inventory management, and quality control. In the manufacturing and printing industries, AI has cut turnaround times, reduced waste, and improved on-time delivery rates (often by 15-35% in reported cases). For broader products, it enables dynamic pricing, demand forecasting, and automated fulfillment.
- Distribution and reach: AI powers recommendation engines, personalized marketing, and logistics optimization (e.g., route planning, micro-fulfillment). Products "find" customers more efficiently, much like when the printing press was first invented, printed pamphlets spread via trade routes. Digital products (software, content, services) can be delivered instantly at near-zero marginal cost.
- Democratization of product creation and markets: Small players or individuals gain the ability to build products that would have historically required larger organizations. A creator can prototype and launch niche-tailored versions of a product without massive capital. This echoes how the printing press empowered reformers, scientists, and merchants. New business models emerge—direct-to-consumer customization, AI-generated product lines, or automated services—while disrupting intermediaries (e.g., traditional design agencies or manual assemblers).
- New industries and job shifts: The printing press gave rise to publishing ecosystems and related trades. AI is spawning new roles in AI oversight, data curation, and ethical governance, while automating routine tasks in product management, testing, and support. Scribes didn't vanish overnight; skills evolved.
A word of caution from the printing press analogy
The printing press was ultimately a hugely positive force in the world, creating great opportunities and leading to the age of enlightenment and the exponential growth of science and knowledge. However, from my research, there are some things we can learn from its introduction.
- Quality, trust, and errors: A flood of AI-generated designs or content risks "hallucinations" or low-value variants, akin to printing errors or propaganda. Also, AI biases lead to mistrust. Delivery systems may need better "proofreading" (validation layers). This blog is an example, of course. I used AI to help me learn about the printing press, but I also read a number of interesting blogs written by historians on the subject. This helped me make informed decisions about the content that I used. It also opened my eyes to lots of things happening in the late 1400s! But that is another story :-)
- Displacement and inequality: Roles in design, manufacturing, planning, or logistics could shrink initially, requiring reskilling (like literacy post-press). Benefits may accrue unevenly to those with access to AI tools. In the early 1500s, the position and power of scribes, town criers, and other roles did change. However, the speed of that change was more gradual. Organizations' adoption of AI will also be slow; the financial opportunity to lay people off might encourage a more rapid adoption of this technology.
- Regulation and ethics: Just as societies developed copyright, censorship debates, and education systems, AI demands frameworks for bias in product recommendations, intellectual property in generated designs, and safe deployment. The current rules on ownership of content and derived content will be under close scrutiny, and because of AI, that scrutiny will be possible, which will, in turn, lead to governance and review being first-class citizens in product delivery processes.
- Pace difference: The printing press's effects unfolded over decades/centuries; AI's are compressing into years, amplifying both opportunities and risks.
The Three Things That Become More Important Than Ever
Ultimately, I am a techno-optimist, believing that technologies such as generative AI and automation will make the world a better place. For example, you might have heard about how Dyno Therapeutics is using AI and Scrum to tackle some of the world's nastiest diseases. This is a story of positive use of this technology, and I am sure there will be many more. The reality of AI is the technology, like the printing press, will impact all our lives in both positive and negative ways. As an agile product developer, I think it will require a focus on three things:
Improved and changed innovation pipeline: Not only do we have the opportunity to try more things because of the technology's impact, but we might now be able to pursue things we would normally avoid due to a lack of information. This means changes to how we do product/feature discovery in our value cycle. There will also be an opportunity to encourage our teams to engage at a higher level of discovery and ask bigger questions about the why, extend value, and really innovate.
Governance becomes very important: Trust and risk at every level of the product delivery cycle have changed as we employ more AI content and automation. This requires us to take a step back and review our governance processes, ensuring we have real transparency and separation of controls. For example, my technology team has implemented a dueling Pull Request (PR) review model in which one AI writes the code, another AI reviews it, the first AI reviews that review, and another AI manages the overall duel and provides transparency. This level of oversight shifted my tech team from doing the work to spending more time managing the governance process and reviewing the agents involved.
Discipline and professionalism: It might not seem odd for an organization that has always emphasized the importance of professionalism in agile delivery to now double down on those ideas. But the motivation for being more professional in Scrum was the increased opportunity and risk that a Scrum-like process can pose to an organization that manages everything in a traditional way. To quote my favorite Marvel comic: "With great power comes great responsibility." Agile was always about empowering teams, and AI has amplified that power exponentially.
It is very easy to use the printing press analogy when talking to people about AI at a dinner party, but the time I took reviewing the ideas and aligning them to product development was interesting and insightful. I also have some historical sources that I did not know before. I encourage everyone to take some time and explore analogies to explore ideas, and AI enables that process, after all, it does have access to the body of all human knowledge :-)