AI Prevails by James R. Simpson

Author:James R. Simpson [Simpson, James R.]
Language: eng
Format: epub
ISBN: 9780578854724
Publisher: James R. Simpson
Published: 2021-09-15T00:00:00+00:00

USE OF THE PRECAUTIONARY PRINCIPLE IN RESEARCH AND DEVELOPMENT CASES

Existential risk from superintelligence development will inevitably manifest itself—quickly in many cases—without sufficient warning for people to take remedial action. As of 2017, over forty organizations worldwide are doing active research on artificial general intelligence (AGI). That refers to the ability to accomplish any cognitive task at least as well as humans, or to “a machine that has the capacity to understand or learn any intellectual task that a human being can.”154 Unfortunately, there are less responsible individuals and miscreants who can attempt to hack into the developers’ work and algorithms to instill their own beliefs. The interest of some, notably singularitarians, is the creation of posthumanity conditions.

In short, considering the above situation, it would seem to be immoral to not give high priority to putting the precautionary principle into force on AGI research in the near term. However, there are two caveats. First, a precautionary principle court case on the topic would have to be created. Second, a working committee would have to be established within a reasonable time frame in a location with statuary power where minimal political interference can be assured.

There are compelling arguments in the Asilomar Principles that risks posed by AI systems, especially catastrophic or existential risks, must be subject to planning and mitigation efforts commensurate with their expected impact. The precautionary principles do include a method to carry out that practice. Although it is not made explicit in all legislation, the EU Commission on the Precautionary Principle states that measures based on the precautionary principle should be periodically reviewed. They should then be amended as necessary when new evidence is available.

Some EU legislation (e.g., Directive 2011/65/EU40) expresses the provisional nature of precautionary measures arising from the precautionary principle and lays down a requirement for review in the light of new evidence, or a requirement to develop evidence. This requirement allows for assessing whether precautionary action has produced the intended consequences. It is also a method to check “whether measures put in place need to be modified, taking into account new information or knowledge that may reduce the degree of scientific uncertainty.”155

Kenisha Garnett and David Parsons reviewed fifteen cases in a study on EU and member states. They found there was a lack of guidance on what conditions justify a reexamination of the potential risks and who would be responsible for producing the evidence required for risk assessment. The conclusion was that, despite guidance not being perfect, there is at least an international system in which there are legal procedures for risk assessment and control.156

The United States is the logical site for a working group, considering the extent of research and development on AI in that country. But there are drawbacks. At the very least, the working committee would have to have statutory authority and wide recognition of the precautionary principle, conditions that the United States is not willing to grant at this time. Considering that legal proceedings will be a lengthy process, if the United States does not give high priority to that requirement, other venues will have to be selected.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.