Object Detection in Dynamicland

Please watch this video or read the comic to understand what this post is about.

Eventually every object could contain a batteryless sensorful computer. Chips the size of RFID tags could have many sensors like LIDAR, Elevation, advanced spatial detectors, even a hygrometer.

Every object could be injected with this smart chip for use in Dynamicland. Each object would be aware of itself. It would know it's own dimensions and the correct orientation of itself. For instance a lamp would know and it could tell you that it is on its side. A useless example but the potential of the internet was also hard to imagine at the earliest stage. A future IKEA bed could illuminate itself (directly or indirectly via communication to your other objects) to help you know how to assemble itself.

These dumb everyday objects would be made smart in a very low level way. They wouldn't connect to Twitter unless you programmed them to but that's not the point. Eventually every chip will be both cheaper than a banana and able to compute more numbers than there are atoms in the universe. Storage will no longer be an issue once we can store information reliably within qudits.

These smart objects would communicate with each other to validate and invalidate truths. For example, a sofa might detect a person but through communication with other smart objects (located nearby but not at the same height) they will conclude that it is more likely that it is a cat not a person.

It might sound bizzare but this is the logical conclusion if we decide to continue in the physical world. This is an eventual reality. All objects that have a UPC will be made smart and even some without a UPC.

Together all objects in the world will be connected in a giant mesh network (if they are close enough to the nearest node). All of these objects will be able to build a virtual representation of reality through interpolation and ‘hive mind’ data. Smart interaction between objects is fundamental to useful-ness of the environment itself.

Bret Victor mentions video cameras but these are not really at all necessary. Each smart chip should be capable of recording and storing all of its past history within the state of a qudit.

https://medium.com/@wittkensis/reflections-on-dynamicland-65158b06196

addendum 3 Aug 2019: I discovered the work of Mark Weiser “ubiquitous computing” and it is probably the direct inspiration for Victor’s Dynamicland.

I think there are several good merits of the idea but I don’t think it will scale beyond affluent communities until the cost of one nano-computer goes down 100x-1000x the current price.

To be truly ubiquitous computing devices will need to be able to materialize and immaterialize themselves in ways that would look like a magic trick to the average person today—but I think it’s definitely possible for these computers to be as small as a few mm of graphene, to carry a battery that lasts a few hundred years, many sensors and storage space, and have an interactive capability at least to interface with other computers.

These nano-computers need not have screens. In fact, the use of screens would make them un-ubiquitous, even audio might not be cool because all the audio and visual interaction could be done from within the brain rather than tainting the shared environment (which might conflict with the preference of someone else). The computers could be sold and distributed in a bag of flour as a dust, or as a paint for example. They would cover the whole environment and record data, a complete history of their life, everything that they have ever seen, heard, and experienced.

The “Disappearing Computer (DC)” approach raises some fear among users because of the increased possibilities of being observed and of loosing control over private information, due to hidden functionality based on embedded invisible devices. This fear is certainly not entirely irrational, and the potential of DC indeed enables invasive capture of private data. On the other hand, as DC is devoted to collaborating with the user and supporting him, it needs to know things about the user. How can this be achieved without the user being put in embarrassing situations where personal data are disclosed to the wrong person, or used by others against his will? As privacy is a cultural notion, it may vary in time and space. The limit between privacy and isolation is not easy to define; as the right to be left alone competes with the right to be informed.