I'll let you check out the talk and slide deck below for the whole thing, but I want to give a taste here of what this theme of my work is all about.
Essentially, my core thesis is that general purpose tools are super productive places to invest in because they have big effects across many industries. People lately see the amazing work that software is doing to transform our world digitally and physically. If we're looking at where specifically to invest though, my take is that the physical engineering spaces are far underinvested in and therefore very high ROI is achievable in a financial and societal sense.
In the beginning of the talk, I start off answering the question: "Why do tools even matter?" by pointing to the (currently unappreciated) contributions that the industrial / scientific revolutions have given us in our quality of life, etc. using historical data. Looking at the data shows that the physical inventions (medical and infrastructure mainly) often have the greatest effects on human outcomes like life expectancy. Further, because physical tools are required for most efforts to really improve human lives, and because they are systemically underinvested in, that there is a huge case to made for accelerating the pace of innovation in hardware tools that will then have a broad impact on hardware innovation across sectors and industries. The question then is: "How do we get better at building hardware?" My answer is essentially to take a look at software engineering because it gets to live in an almost post-scarcity world where time and money aren't factors in most areas engineering productivity. Therefore, software tools and process have already found some of the abstractions that can be repurposed to improve other hardware engineering disciplines. One of the issues in the hardware engineering tools space though is that there are only 4 big companies making CAD, PLM, and related products, and those systems have lock-in effects akin to ERP or similar software systems that mean that those firms have little threat of customer's switching to new CAD tools, which further restricts innovation. Compare this to the open source world that dominates software where engineers use free tools that they also can improve themselves, and you get the ultimate in feedback loops where the open and closed sources tools all have a place and there is a high level of competition, innovation, and investment in those toolsets because everyone recognizes the leverage those tools give engineers. By contrast, the lack of open source development tools in the hardware community occurs because mechanical engineers, or even most software engineers, can't realistically create/contribute to the highly complex CAD programs that they use. The "big 4" CAD companies then have high switching costs and new barriers due to their lock-in and high initial investment to create CAD tools. Even efforts like OnShape (innovative cloud based CAD) that were started by the founders of the most successful CAD software (Solidworks) was unable to get widespread adoption or achieve broad enough feature parity due to the above. The net effect is that mechanical CAD is largely the same as it was 20 years ago and the leverage it provides engineers negligible compared to the same advances in software tools throughout even the past 5-10 years, let alone 20. Another part I cover is the stagnation of factories and their adoption of automation technologies. Essentially, they experience another set of issues coming from (1) the commodification of their inputs leading to underinvestment in new tech, (2) inability to find / train talent to implement those technologies, (3) path dependancies of equipment purchases, and (4) a general lack of profits that does not attract enough investment and high quality talent into the sector. In the past 10 years, much hype has occurred around 3D printing technologies that promise to end run all of that, but as I show in the talk, manufacturing processes are more like an ecology of species specialized to their particular niche to achieve a given level of performance and cost. Therefore, instead of 3D printing saving us, we must instead save all the other processes by digitalizing them to make them faster/cheaper to setup (ie program / retool the machines) and more available via modern software UX, as opposed to phone calls.
Towards the end of my talk, I explore to emerging technologies in engineering tools and imagine what the future world of hardware engineering tools could look like after adequate investment allowed it to multiple engineers' powers of creation by orders of magnitude, allowing hardware innovation to go exponential like software has.
I welcome any questions you have on this, so please hit me up on Twitter: @NickPinkston
"The Future of Tools" at What's Now New York:
Slide Deck from the Talk:
Comments