B. Assessing Novelty/Innovativeness | Privacy International


Public-private partnerships might very well involve a technology that is well-known and already widely deployed in other contexts, but it might also be a ground for innovation and novelty.

We can identify two main types of innovation:

  1. A new technology that isn’t widely used or hasn’t been deployed in real world context (outside of a lab or a research paper);
  2. A new set of features added to an existing technology which greatly expand its performances and capabilities.

Another type of “innovation” may involve deploying existing technologies in new contexts. Assessing these deployments will mostly require identifying governance concerns (Section 4 below).

i. Entirely new technology

In the case of an entirely new technology, the novelty factor is usually obvious as the technology at stake is likely to be little known. New breaking-ground technologies are rare and pose several risks as they haven’t necessarily been properly tested or might have unexpected side effects. Approaching an entirely new technology is hard for external actors and it might be difficult to gather relevant information. Nonetheless, there are several common risks that come with innovation and might be worth exploring:

  • Disparities between the test environment and the real world impacting the technology efficiency
  • Insufficient testing meaning the population it’s deployed against is effectively beta testing it
  • Overestimation of the technology capabilities and accuracy – the technology doesn’t deliver, produces too many errors
  • Creation of new problems has been overlooked – this might come from the cost of maintenance, the sustainability of the project in time, the issues arising from edge cases, or the public body being trapped in a contract that can’t be fulfilled by anyone else
  • Function creep, the technology can be used for more purposes than it is initially intended for
  • Lack of transparency/accountability if the technology is protected by trade secrets and/or under proprietary licencing

ii. New features/capabilities

In the case of an existing technology with new features or capabilities, the innovation might be more difficult to spot but can have an important impact on the use made of the technology. New features and capabilities might come from a technological leap at the hardware or software level. This could be for example with new Computer Processing Units (CPUs) that are much more powerful than the previous generation, or novel technical developments such as quantum computing. For the software side it can be the development of new processing techniques such as the rise of deep learning and equivalent AI solutions.

These innovations can also simply be the addition of an existing technology onto a solution, for example by mounting Radion Frequency sensors to a multitude of small satellites (such as CubeSat), an innovation made possible by the cheap price of these satellites. The risks that arise with the addition of new features or capabilities are more specific and should be easier to identify. Here are some of the risks that arise with such innovations:

  • Added capabilities might be unnecessary for the technology to perform its initial function (e.g. equipping body worn cameras with temperature sensors)
  • Added capability hasn’t been properly tested for the environment it’s being deployed in and might yield inappropriate results (e.g. deploying a neural network algorithm for the judiciary system)
  • New feature/capability gives the technology a far more intrusive reach (e.g. improving image quality of video surveillance cameras)
  • New feature/capability makes the technology much more efficient and enables mass application (e.g. mass interception and processing of internet data)

iii. A note about technical protocols and standards

As mentioned under the data transmission system section, protocols and standards might be an interesting thing to look at and understand when trying to take apart a technology. It can give information about whether the technology deployment is taking place in an already developed and standardised environment or trying to define new standards. Here are some useful definitions and standardisation bodies that might be relevant:

Definitions:

  • Protocol: a protocol is an agreed language that allows different elements to communicate. One of the most well-known protocol is the internet protocol suite also known as TCP/IP. Protocols are usually standardised and answer to a specific set of rules. They allow any new player in a market to easily develop a product that will be able to use existing infrastructure and communicate with other products. For example, with TCP/IP, anyone can create an internet connected device that will communicate with a server or similar devices around the world.
  • Technical Standard: Norm or requirement for a technical task to be operated. Technical standards are more abstract than protocols in that they don’t offer hard defined rules for a given programming language or technology. They establish uniform principles, methods and processes that should be followed when developing a technology. The goal is to ensure interoperability between devices and systems (e.g. making sure an external hard drive from a company other than your computer manufacturer will function in any computer). Standards can be developed privately or unilaterally by standardisation organisations. Example: Universal Serial Bus (USB).

Standardisation bodies:

  • International Telecommunication Union (ITU): the ITU is a UN organisation in charge of radio communication and standardisation. It works to ensure that countries and private actors agree on standards and protocols to avoid collision and promote development. Standards developed by the ITU are referred to as Recommendations. Some examples of what it does:
    • Manage the radio frequency spectrum (defines which part of the spectrum can be used for what, and by whom, e.g. Wifi and Bluetooth operate between 2400 and 2500MHz);
    • Develop and maintain the Open Document Architecture, an example of a free and open-source standard document file format that any software developer can use for word processing;
    • Publish recommendations regarding armouring of cables to limit interference;
    • Working Groups develop recommendations on topics such as Quantum Information Technologies for Networks and Artificial Intelligence for Assisted and Autonomous Driving.
  • Internet Engineering Task Force (IETF): an open standards organization, which develops and promotes voluntary Internet standards, in particular the standards that comprise the Internet protocol suite (TCP/IP).
  • International Organization for Standardization (ISO): a standard-setting body composed of representatives from various national standards organizations that publishes worldwide technical, industrial, and commercial standards. Example of standard: ISO 80601 that ensures that thermometers are calibrated the same way in different hospitals.
  • W3C: standards organization for the World Wide Web.

Standardisation bodies who have open discussions about standards are interesting (although often hard to penetrate) places to look for lobbying and influence. Private companies or countries might use them as gateways to push a technical solution which might have political consequences. For example, in 2021, the W3C has an “improving web advertising” business group in which Google suggested a replacement to third party cookies that still enable tracking and targeting.



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *