Data is eating software.
I believe that software is increasingly just a tool for collecting, managing and delivering (raw and processed) data to people and other machines. Companies and people need to focus more on proprietary and continuously updated/refreshed data as THE PRIMARY ASSET and treat software (even modern AI) as the necessary tooling to build, maintaining and applying data as a strategic competitive asset.
I was recently compelled to go back and reread and reflect on Marc Andreesen’s excellent 2011 “Software is eating the world” post. It caused me to consider not only the current reality check we are experiencing with the latest AI hype cycle(this is my third;) but also the broader role of software relative to AI and data. Marc makes many great points and since 2011, software has arguably been the core driver of many productivity improvements in our economy. AI is now driving another level of productivity improvement in software engineering especially. Software has also primarily enabled non incremental capabilities — in both B2B and B2C context — that seemed like scifi only a decades ago — driver-less electric cars, conversational AI, missions to Mars etc.
However as I look forward couldn’t help but realize the increasingly important role of data and that software’s primary role going forward is to be a tool for the creation, curation and strategic application of data within organizations and for people — both consumer as well as B2B. Software is increasingly more of a commodity and data is increasingly the unique and differentiating asset.
Data as strategic competitive asset is already a core assumption within most consumer businesses that are digitizing at a blistering pace. Many of the largest consumer internet businesses are already pushing the boundaries on what is possible and what is required to manage and apply data as a strategic competitive asset. In my opinion, relative to consumer businesses, the data capability/sophistication gap in B2B/industrial/government organizations is much larger (with a few exceptions). In my experience, most non-consumer companies/orgs have only just begun to get their feet underneath them with regards to software but haven’t yet begun to prioritize data as the primary strategic competitive advantage. Now is the time for many of those companies to double down on data as the core focus and start thinking of software a necessary tool to develop (+ responsibly manage/govern) the real asset — data about their entire business — customers, products, suppliers, competitors, purchases, employees et al. Along with many others I recently posted about how data can/should be managed as product here which embodies many of the core principles that I’ve seen work well.
I increasingly believe that we’ve done much of what is possible with software — certainly AI will give us another bump in terms of productivity, accessibility and capability — especially wrt writing code. Especially in context of the impact that AI will have on how code is written with Copilot et al — I believe that we are at a point where data has become more important than software. Certainly data is the primary unique differentiation for most AI systems — the more and better your data — the more likely you are to have better and more useful AI models and user interactions. I also believe that in order for AI enabled systems to broaden their context, more and better data will be the proverbial long pole in the tent. One of my favorite examples of this at the moment is how modern AI systems are beginning to embrace RAGs and knowledge graphs as tools to add contextual nouns to conversational AI — good recent reference here.
I firmly believe that much as Marc Andreesen validated the role of software in 2011 that we’re at a pivotal moment to recognize that data is THE thing for the coming decades. The companies that have more well managed proprietary data and know how to apply the data strategically to their businesses will become as dominant over the next 20 years as software savvy companies have become in the past 2–3 decades. Many of the best tech companies in the world already operate with data as a core strategic asset and those companies will have a disproportionate advantage over non-data savvy companies going forward.