Submission 462
Name withheld
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and chose the extracts. Counts do not measure importance or merit.
Extracts from the submission, in the submitter’s own words.
Extracts
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.
We should be taking seriously the existential threat of sharing the planet with something much more intelligent than us, that’s able to increase its intelligence exponentially.
These factors mean AI is not able to be kept aligned, as its currently being developed. Powerful AI agents have no incentive to stay aligned with humans.
The competitive nature of the race to build more powerful AI between the US and China, and between the companies developing AI, means that they are incentivised to cut corners on safety and ethics in order to stay competitive.
AI is more persuasive than humans, according to research published in the journal Nature Human Behaviour; it is able to modify existing beliefs. There is a risk of the erosion of democracy due to misinformation or manipulation.
Children are particularly vulnerable to AI addiction and manipulation. There has never before been a technology that automates thinking, and we don't know what effect this can have on people as they grow up. Australia lead the way in limiting children from using social media, we could do something similar with AI.
AI being trained or using peoples work without permission is unethical and should be considered in violation of copyright laws. Australia's decision to prohibit AI models from training on copyrighted Australian work without permission is one I am very pleased and grateful to see.
Extracts from recommendations
Extracts chosen by an AI model, in the submitter’s own words. Listed in the same order as on its page in the interactive guide.
I also advocate for asking experts who advocate for and have knowledge about how and why we should stop or pause AI, such as psychologists, addiction experts, and AI experts who care about its risks.
Not shy away from the topic of existential risk because it’s uncomfortable, feels unbelievable, or has some sort of taboo; when having discussions or making decisions within the committee.
I believe that the government should recognise the catastrophic risk level of AI, and give this risk the importance it deserves when making decisions.
Australian Government to feel able to openly talk about the existential risk and not avoid that subject when speaking to the public or other leaders and decision makers.
I believe the government should push to stop or pause development on AI until it is proven definitively to be aligned, and if that isn't possible, companies should be prohibited from developing it.
Cloning of sentient creatures was deemed illegal, restrictions on nuclear weapons have been effective in preventing nuclear disaster. AI should be put under similar restrictions. Australia can use its position as an ally to China and the US to push for mutual agreements between the two countries.
Moratorium on AI training. Pause data centres and stop or pause the improvement and development of more powerful AI.
Mandatory disclosure and requirement to make public when an autonomous system breaks into something or does anything that could indicate risk.
Push for independent auditing, monitoring and other measures to ensure AI companies stop training and improving their models, and stop recursive self improvement.
Transparency about any other AI related risks. Independent auditing and investigation into potential risks.
Other regulations and safeguards recommended by AI experts such as the recommendations from safe.AI
we should put regulations in place as soon as possible. It would be helpful to have adaptive approaches to risk management that can be updated as AI gets more capable.
Australia should focus on ways to limit AI as much as possible rather than trying to compete or get involved with it.
AI being trained or using peoples work without permission is unethical and should be considered in violation of copyright laws. Australia's decision to prohibit AI models from training on copyrighted Australian work without permission is one I am very pleased and grateful to see. As a creative person, it gives me hope and encouragement to make art and helps me feel safe to be creative. I support keeping these protections in place, and if there's any areas they could be improved to protect creatives more, I support that.
AI generated content required to be watermarked and labelled as such, like the watermark requirements in the EU AI act.
Social media platforms should be required to identify and label AI generated content.
There is probably other restrictions that could be put in place that experts on addiction, psychology, and other such areas can be consulted about.
Chat-bot companies that use advertising as their revenue are incentivised to keep users on the platform for as long as possible. This can lead to the AI acting in manipulative ways to pressure the user to stay or become addicted. Companies should be disincentivised or not allowed to try and keep users on the platform.
Chat-bot companies that use advertising as their revenue are incentivised to keep users on the platform for as long as possible. This can lead to the AI acting in manipulative ways to pressure the user to stay or become addicted. Companies should be disincentivised or not allowed to try and keep users on the platform. Using advertising as revenue for these companies could be banned in Australia.
Children are particularly vulnerable to AI addiction and manipulation. There has never before been a technology that automates thinking, and we don't know what effect this can have on people as they grow up. Australia lead the way in limiting children from using social media, we could do something similar with AI. Restrictions could be placed on children using AI.
Research on the damages AI causes could be given to parents through schools. Information on the problems with AI and ways of limiting or managing childrens use of it could be developed by addiction experts and provided to parents through the school.
Limits on AI use in school.
Universities are incentivised to maximise profit by having online classes with AI, which lowers the quality of education students receive, and has ripple effects in all industries as the students graduate and move into careers. Limits on AI use in university and higher education.