There are several software based paradigm shifts happening at the same time.
- Most of the business world has yet to transition to social and mobile media. Some businesses have yet to completely transition their paper records to electronic records and utilize their aggregate value (dirty data, weak data, data mashups, data visualization tools for utility/value discovery, task automation).
- Business desktop tools are not anticipatory. Automation tools for routine bureaucratic business tasks aren’t here yet. Use of pervasive desktop aggregation and big data (dirty data and weak data) has yet to come online. Automated multi-task scheduling capabilities (such as the Gantt charts used by project managers) have yet to be adapted for personal use and diffused to the greater working population. Aggregated bottom up multi-tasking traffic optimisation across social networks has yet to be developed.
- Pervasive and ubiquitous anticipatory AR OS systems are an opportunity ready to be exploited. Most cameras are dumb right now, and there is low fluidity in video network capability and data aggregation. RFID and the internet of things is just sitting in a box waiting to be unpacked. Every commercial vehicle can be automated from trucking to mining equipment, construction with quadcopter swarms to warehouse drones.
Most tech giants are aware that there just aren’t enough software engineers worldwide, or in the pipeline. Even with Viet Nam including a sophisticated software engineering curriculum in their secondary school education.
So what are the big tech giants doing about it? What should they be doing about it? The most immediate thing they are doing is outsourcing to the existing worldwide workforce. This is a bit turbulent in a global recession, and brings us back to the ComputerWorld story above. Software Robotics
The most obvious solution to the software engineer shortage is to develop a new software engineering tool set. IMO this is the biggest challenge/opportunity in software engineering right now. Software development needs to become easy enough that it can be done by semi-skilled labor.
Unless software is written in machine language, code is meant to be read by other humans. Many of the new scripting languages brag about “semantic sugar.” It’s the idea is that when a programming language is closer to human spoken language, there is increased fluency and increased productivity. These upper level scripting languages are a move in the right direction, but instead of semantic sugar there needs to be a semantic happy meal. Think of the command line OS tools from the 80s (like DOS) compared with SIRI and GUI touch screen that can be used by toddlers or the illiterate.
Intuitive human friendly tools for software engineering will increase not only the worker base it will increase the entire user base. Putting DRY (Do-not-Repeat-Yourself) automation tools in the hands of individual workers will explode productivity and grow the economy in ways that cannot yet be foreseen. It will be a paradigm shift as powerful as popular literacy and numeracy.
As an example, here is an MIT coding toolset for kids. What is the adult version for the workplace?
The Samsung Intel backed app called MindMeld can listen to a conversation of up to eight people at once and creates a dynamic search engine results feed.
Mindmeld is being targeted at networked chat. However the same search feed technology is well suited to a pervasive ubiquitous paradigm, as user experience flows between glass, phone, tablet, laptop, browser, and etc. A pervasive record of device use (desktop, browser, phone, text, geo) over hours, days and years allows search engines to find and use hyper personalized (and valuable) preference patterns in deep time. Again, it’s worth emphasizing that the search results would be a dynamic feed rather than a static results page. So the context of the users situation and (forward/backward) patterns in time would constantly be aggregated into the feed. What is the user browsing now? What email conversations are they having? Where are they?
Socially mediated search – social dynamics and information patterns from parallel (pervasive) records are also important. One major paradigm change in networked data is the mash-up. When feeds collide amazing things happen. Humans are social animals. When rich social pathways are used to direct data collision, value is created. Exposing parallel search feeds across the network (and exposing contexts in search feed information pathways) allows for social search grooming and re-versioning.
This presentation on MindMeld technology and mapping techniques can be applied to visualization tools and dynamic data.
Thermodynamics is awesome! If you want to know how the universe works, thermodynamics is an essential piece of the puzzle.
Check this Harvard AI software that applies entropy to planning. Entropy as intelligence?
For another eye opener – check this sick video on autocatalysis and think about the business cycle, forest fires and other natural systems.
The Defense Advanced Research Projects Agency (“DARPA”) is one of the few tech programs left in the U.S. Fed Budget. This isn’t surprising. The history of technology is strongly associated with war.
So like, what are they up to?
DARPA thinks Moore’s law is moving too slow. They are funding research into new material science, new sensors, new transducers and Quantum engineered devices. Awesome! Check the details at Next Big Future.