Submission 398
Ms Wendy John
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 journalist and parent, who discloses using ChatGPT to help structure the submission, says they are excited by AI's potential but concerned about catastrophic risk. Argues that low-probability catastrophic harms warrant action, and calls for safeguards, human control, expert assessment mechanisms, diverse advisers and international cooperation on testing and safety standards.
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 where the potential consequence is catastrophic action is warranted even if the risk is low, which the author says nobody can reliably quantify.
Nobody can reliably quantify that risk. Yet where the potential consequence is catastrophic, even if the risk is low, we must act; any Workplace Health and Safety officer will tell you that.
Argues that Australia should establish mechanisms for expert assessment of machine sentience rather than assuming machines are or could never become conscious.
From an ethical lens, machine sentience or morally relevant forms of artificial consciousness emerges in the future, our regulatory framework should be capable of responding. Australia should establish mechanisms for expert assessment rather than assuming either that machines are conscious or that they could never become so.
Recommends placing independent, ethical and compassionate thinkers at the centre of national AI work.
I urge Parliament to put our smartest and most independent minds at the centre of this work; our most compassionate and ethical thinkers as well.
Says Australia has an opportunity to lead international cooperation on testing, transparency, safety standards and governance, since AI risks do not respect borders.
AI does not respect national borders, and neither will many of its risks. Australia has an opportunity to lead in building international cooperation around testing, transparency, safety standards and governance.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Develop solutions on safeguards needed if AI becomes substantially more autonomous and capable, how humans retain meaningful control, and how to prevent deployment before risks are adequately understood.
Australia should be creating solutions around these concerns: - what safeguards will be required if AI systems become substantially more autonomous and capable; - how humans retain meaningful control; - how we prevent systems from being deployed before their risks are adequately understood.
Establish mechanisms for expert assessment of whether machines are or could become conscious, so the regulatory framework can respond if machine sentience emerges.
From an ethical lens, machine sentience or morally relevant forms of artificial consciousness emerges in the future, our regulatory framework should be capable of responding. Australia should establish mechanisms for expert assessment rather than assuming either that machines are conscious or that they could never become so.
Guard against exploitation of new forms of technology wherever compelling evidence warrants moral consideration.
Our principles should guard against exploitation of new forms of technology wherever compelling evidence eventually tells us moral consideration is warranted.
Put the most independent, compassionate and ethical thinkers at the centre of national AI work.
I urge Parliament to put our smartest and most independent minds at the centre of this work; our most compassionate and ethical thinkers as well.
Involve computer scientists, safety researchers, ethicists, faith ministers, engineers, cybersecurity and national-security experts, behavioural economists and community representatives in national AI decision-making.
Australia needs people with this kind of approach involved in national AI decision-making: computer scientists, safety researchers, ethicists, ministers of faiths (try Reverend Catherine Solomon of the Uniting Church of Australia, QLD), engineers, cybersecurity and national-security experts, behavioural economist and representatives of the broader community.
Include deep thinkers, philosophers, artists and ‘court jesters’ in probing overlooked issues, and ask children.
Pull in the deep thinkers and philosophers, the artists and ‘court jesters’ to probe what we haven’t yet considered properly. Ask the children.
Lead in building international cooperation on AI testing, transparency, safety standards and governance.
Australia has an opportunity to lead in building international cooperation around testing, transparency, safety standards and governance.
Treat advanced AI as requiring unprecedented cooperation and remove barriers between disciplines, agencies and nations.
I ask parliamentarians across political divisions to treat advanced AI as an issue requiring unprecedented cooperation and to remove bureaucratic barriers between disciplines, agencies and nations.
Do not wait for catastrophe to establish the institutions and regulations that may be needed.
Please do not wait for catastrophe to establish the institutions and regulations we already know we may need.