“As an economist, I understand productivity; people lose jobs when technology changes. But the valuations of these AI companies are literally based on the assumption that hundreds of millions of people will lose their jobs and employers will buy AI services instead of labour services.”
Dr Denniss said that while much of the attention was focused on the technology itself, he was more concerned about the broader economic consequences.
“What is the ATO’s plan? What are the state government plans?” he asked the committee.
“Again, all that money that Anthropic is betting on is coming from somewhere, and that’s going to have very real consequences for individuals, for communities, for the economy, for the way Commonwealth budgets are drawn up.”
Dr Denniss also challenged the argument that Australia risked being left behind unless it rapidly embraced AI, warning that national security concerns should not be used to justify an unrestricted rollout.
Sure was, such a shame regurgitating the internet doesn't get you to self improving AI. LLMs are pretty much tapped out in advancement (logarithmic wall). Turns out to be reasonably useful for things where you can quickly test if the answer is correct and not one of the 'hallucinations', like mathematics and programming.
Newsflash, 95+% of jobs are not quickly testable for correctness and require judgement, not guessing the next word. Some bullshit paperwork jobs may go, but we can count ourselves lucky these evil pricks didn't get there, this time. The societal destruction would be immense. Heed the warning.
Now it's a circular money Ponzi scheme where seven companies see theoretical productivity gains and the rest of the US economy is much the same as it was.
Australia following the US economy down the incoming crash from this scam is something where we need to minimize the harm, not jump in.
I think the unfortunate reality is a lot of office jobs are repetition and busywork. anywhere that has extensive SOP's are a target for reduced work due to automation. Even front line jobs like managing a restaurant have good whack of paperwork/administration that could be done by a program, next minute 10 restaurants are covered by 1 manager and each has a lower paid 'team leader' handing spot fires and discipline.
AI wont do everything but if the costs can be kept under control, or used for one time expenditure to write programs to automate tasks then it will cause permanent layoffs, (not the kneejerk layoffs driven by AGI promises).
This is arguably worse as if everyone got laid off then there would automatically be a cultural reaction and real change would have to be made. If there remains some 'good' jobs then we are right back where we are now, defining someone's worth by how much they earn and looking down on those who don't.
You're not wrong, it's a crying shame that increased productivity decoupled from the worker's income. Now it goes straight into the 1%'s pocket. Even worse that the dream of an ever lower hours working week with automation got squashed. Shit, it's gone into reverse, for no other reason than to make a number in an account go up. Those BS jobs should go, but they're so linked to survival that it will do massive damage to society, not least by making people ever more desperate for the crumbs of jobs remaining.
Yup, although it's more by how much wealth they have IMO, which is worse, especially when morality gets tied in.
Agreed, but they will inevitably circle around to trying again (and again if necessary) given the chance, hence 'Heed the warning'. It's (perhaps) just one significant advance in AI theory away. It would behoove us to lay the groundwork, build the community, and do the thinking so that when the real change has a chance it's not a kneejerk away from the current hellscape, but a deliberate change toward a coherent positive. We've had our neoliberal two steps back, time for three steps forward.