The Roadmap: 6 Lessons for Navigating in the AI Era
The Roadmap: 6 Lessons for Navigating in the AI Era
As the company offsite comes to an end, the CEO captures the takeaways in this last post.
If you're just joining us, this series follows the journey of a brilliant product manager, you. You've steered the product strategy of a company, Warehouse Software Inc, through 3 decades of technology disruptions. But in 2026, the company faces its biggest test yet - the ascendance of AI. The CEO has called an offsite, and the team is grappling with the implication of AI and how the flagship product, Warehouse Tracker, should respond to the change.
Here are links to the rest of the series:
Moments That Modernized: Four Decades of Tech's Tectonic Shifts
The Economics of Abundance: What Do You Charge When Software is Free to Produce?
Reinvent or Relinquish: You Can't Sell Horseshoes to People Who Don't Own Horses
The CEO strides to the front of the conference room and calls the meeting to order. Though the past two days have been intense, he can see the team is ready to return to their normal routines. Email checks and call breaks are becoming more frequent. Chatter from the office Slack channel is seeping into the discussions.
He understands the desire to get back to business as usual. Offsites untether teams from routine but routine calls them back. These strategic sessions are critical but tiring. Still, he fears the urgency to chart a course through the churning waters of change will fade once they return to the office. Such are the challenges of running an established company. Startups have no such luxury—their survival depends on figuring it out.
The CEO begins, "I'd like to thank you all for taking the time to be here and be present. While it may not seem like it, the conversation here will set in motion changes that will transpire over the coming quarters and years. We've navigated changes before, and I have no doubt we'll do the same here. I want to take this last session to recap the takeaways."
Lesson 1: The AI Snake Eating Its Tail: How Software is Consuming Itself
Over the past 30 years, the tech industry has been transforming the economy. Back in the 1980s, the tech sector accounted for 0.8% of the GDP and 2.8% of employment. Today, the IT sector makes up 10% of the economy but still only 7% of jobs. Tech's growth in contribution to our GDP has gone up 12,500%, but its employee base has only grown 250%.
This demonstrates the formidable job-destroying power of automation. As the tech sector produces a larger share of the economy, it requires fewer jobs. While these tech jobs pay higher wages—85% more than the average US worker—there are far fewer of these jobs than the ones they displace.
Now the power of AI will unleash automation on the software industry itself. Many of the jobs in tech that are necessary to run the great software machinery will be done by AI itself.
We can already see examples of this in how AI chatbots are replacing the need for human-created help documentation, tutorials, and guides. Users no longer have to read and comprehend content put together by people. They can simply ask the bot their question.
AI will automate software testing, monitoring system performance, detecting bugs and security issues, and deploying updates.
Lesson 2: Hyper-Personalization: Scaling Unique Experiences
The holy grail for any business is providing personalized products at a low cost but at scale. This has always been challenging in the world of atoms. Companies segment customers into broad categories to offer tailored options, like minivans for soccer moms and sports cars for SINKs ( single income no kids ).
In the digital realm, however, real-time algorithms can recommend the ideal product for each customer. Today, AI-powered travel sites can generate customized itineraries as unique as the traveler. This revolution is coming to content and software.
In music, for example, we're going from selecting genres to specifying the exact music we want to hear. Want a podcast that conjures the wit of Oscar Wilde to satirize your daily commute? or a pop song in the voice of Taylor Swift explaining string theory? AI systems can compose and produce songs and scores tailored to our tastes. In a sense, AI is giving each of us our own digital genie—magic lamp not included.
And the AI genies are just getting started. They'll continue gaining knowledge and skills to build personalized worlds in every digital domain. The future is cacophonous, idiosyncratic, and intensely customized. One-size-fits-all is over. The age of the algorithmic artisan—the AI genie—is here.
Lesson 3: The Great AI Unbundling: How Software is Breaking Apart and Reassembling
"There are only two ways to make money in business: One is to bundle; the other is to unbundle."
Erik Torenberg
Tech analyst Ben Thompson has explored how the internet enables the bundling and unbundling of products into new forms. What were once bundled into single packages are now unpackaged into components and repackaged as customized offerings.
Newspapers, for example, once bundled news reports, opinion columns, and crosswords into a single product on our doorsteps each morning. Today, those components have been unbundled. Newsletters on Substack deliver analysis and commentary on specific topics. Crossword apps serve puzzles on demand.
The music album has met a similar fate. Unbundled into individual songs, music is now rebundled into shareable Spotify playlists tailored to personal tastes.
Today's enterprise software bundles core functions, data management, and workflows into one application. The decision on how to package the functionality rests with product managers and UX designers, who play roles similar to newspaper editors and DJs. But their days are numbered. While we can hardly conceive of user interfaces and experiences, separate from the tools themselves. That will soon change.
The rise of APIs has enabled a modular, mix-and-match approach to software. Though promising, assembling APIs into customized solutions has required customers to invest in an engineering team or hire consultants—until now. AI-generated software unlocks personalized experiences without the hassle or expense.
The days of one-size-fits-all products are numbered. Competitive advantage will come not from specific applications or features but the ability to combine modular parts uniquely for each client or user base.
Lesson 4: The Customer Revolution: How AI is Upending Expectations and Industries
At Warehouse Software Inc, we support our customers trying to build successful businesses off the e-commerce gold rush. We make the digital equivalents of gold pans and sluice boxes for these modern-day argonauts.
We must stay attuned to customers' changing needs, as many we serve may be impacted by automation. While we're used to our customer's push for automation, we've always built software to be used by humans. Likely, many of the roles we serve today may not exist or be replaced by AI agents or bots.
As we saw in our discussion on the evolution of digital cameras, it's easy to assume that any new technology will be an asset to us. It could just as well spell the end of our path.
Lesson 5: The Coming Human-Machine Blend: How AI Will Fuse Flesh and Code
Technological and social change happens gradually, not instantly. Innovations are adopted slowly, replacing older systems and behaviors over time. This process is often slow, uneven, and messy. Emerging technologies do not promptly displace the old; multiple systems and ways of doing things coexist, sometimes for decades.
Consider credit card imprinters, also known as ZipZaps or Knuckle Busters. To use one, you would insert a credit card into the bed and layer carbon paper forms on top. Sliding the bar across copied the card number onto the forms. Though first introduced in the 1950s, some stores deployed imprinters into the 1990s.
Even today, in the 21st century, some credit cards still have embossed numbering—a relic of imprinters—while also containing magnetic strips and chips. According to the U.S. Payments Forum, embossed card numbers are still the backup mechanism for card processing if a magnetic strip or chip cannot be read, though they may eventually disappear.
As much as we'd like a clean slate reboot with AI, it's unlikely. We will upgrade tech piecemeal, keeping the old around. What this could mean practically is that in the beginning, our dock scheduling software may include a bot or AI agent that calls drivers to coordinate the time. And in time, the agent will call a self-driving truck equipped with its own bot, which it needs to talk to the warehouses that still employ humans to coordinate dock schedules.
For a peek back, check out Google's Duplex Demo in 2018.
Lesson 6: The Perils of Falling Behind: How Companies That Miss the AI Wave Will Perish and Pave the Way for Disruptors
Not all companies successfully navigate major technology shifts. Some, like Noritsu and Columbia Motors, faded into history. Others, like Ford, capitalize on changes to gain an advantage.
When Henry Ford harnessed electrification and the assembly line, he transformed automaking. Many modern companies will miss opportunities from AI, failing to recognize its potential for strategic benefit. This could open doors for disruptors with innovative business models.
For example, warehouse software providers may need to decide between remaining toolmakers or becoming "kingmakers"—partnering with and enabling a new breed of highly automated, large-scale warehousing companies.
Success will depend on managing transitions, not fighting or fleeing from them. The future is there for those with vision to see it—and the patience to get there, one step at a time.
The CEO ended by saying, "In the meantime, I recommend all of you subscribe to the boxcars.ai newsletter."