Submission 456
Ms Gwen Bairstow
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
A private citizen raises four issues: human-led AI development, kill switches, transparency and algorithms. Recommends an international safeguards group, mandatory labelling of AI with standards and a registration list, and an option to turn algorithms off. Says AI has capacity to improve things but also to make them worse if not controlled.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Argues that AI development should be led by humans for the benefit of humans and the natural world.
It really is essential that AI development is led by humans for the benefit of humans and the natural world.
Says autonomous machines are being created without kill switches and that an international approach is needed urgently.
In the meantime, we are creating autonomous, drones, robots and copies of humans that don’t have kill switches. This shows that the risk of AI is already an international issue and an international approach is needed urgently.
Argues that AI must be labelled so people can tell fakes from real things, and that the risk comes from those not doing so.
To ensure transparency, so that people can tell the difference between fakes and real things, AI must be labelled as such. Good companies and individuals are doing this, but it is the ones which are not good which are the risk and which are not doing it, but rather using it for very negative purposes.
Argues users should be able to choose no algorithm at all.
There should be a choice for NO algorithm at all.
Argues that AI will intensify existing profiling by big tech, reducing individual agency and disadvantaging poorer consumers.
Supercharging these examples via AI will reduce individual agency and autonomy to a greater level and many people may not even understand this. One of the many ramifications of this is that the poor will be discriminated against (again) paying the maximum that the supermarkets can get out of them and their discretionary income will be reduced.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Do not dismiss requests from big tech figures for safeguards and a slower pace; heed their warnings urgently.
Please don’t dismiss the requests from the big tech guys who are asking you to put safeguards in place and to slow it down as some underlying ploy to make more money... That would be silly as they are going make more money than they could ever want, either way. Please heed their warnings urgently.
Articulate a vision, in terms an average person can understand, of how AI is going to help people.
Could you please articulate a vision and make it clear as to how AI is going to help us, in terms that an average person can understand?
Convene an international group including all countries, independent of the UN but including its voice, to set measurable, accountable safeguards including an approach to kill switches and monitoring.
Could you please convene an international group that understands these issues urgently, to make a set of safeguards, that are measurable and accountable and include an appropriate approach to kill switches and monitoring of them? This should include all countries and be independent of the UN, but also include it’s voice. Please fix this urgently.
Mandate automatic labelling of all AI.
Could you please mandate that all AI be automatically labelled?
Create company standards understandable to an average person and a consumer-searchable registration list like an improved ASIC list; require human approval rather than automated approval.
Also please have a system which includes a set of standards (that can be understood by an average person) that companies must adhere to and a registration list (like ASIC). Don’t use automated systems to approve these. Keep humans in the loop. The registration list could be searched on (use ASIC’s list but make it better) by consumers so that they know they are dealing with a legitimate organisation, in Australia at least.
Implement age-appropriate internet safety education at every education level, starting in pre-school.
Could you please implement education about how to use the internet safely at all levels of the education system in age appropriate ways, starting in pre-school?
Mandate big tech algorithm-removal options and alternative information-finding methods, no reading of device information, and responsibility for technology through short, reasonable Terms and Conditions Agreements.
Also please mandate that big tech do the following: e Provide the option to completely take off algorithms and provide other ways of finding useful information in the internet; e Do not read information in people’s devices at all; e Take responsibility for the technology they have created by having short and reasonable Terms and Conditions Agreements.