I read the NYT article with interest and am struck with the parallels with machine translation and post-editing. All those "AI audit"-type jobs - if it follows the languages profession, I would put good money on those practitioners being paid at half the rate for AI auditing as they are today for doing the work the AI is expected to do, even though the responsibility is the same or greater.
Post-editing a machine translation requires the same skills as professional human translation plus in addition post-editing skills, an even greater eye for detail and concentration (especially because of the plausibility), and often takes the same amount of time as a human translation from scratch - and sometimes a human translation is actually quicker. Yet rates for post-editing are 50-60% of translation rates if you're lucky. And that's before you even consider the sameness of AI-slop - every contract, every report, every bit of code, all the same...
The languages profession is losing skilled people because they can earn more at Lidl (cf the evidence to the HoL committee on public sector interpreting). If the same practice happens across the "AI-audit" field, who is going to be willing to do the jobs with accountability?
One wonders if AI is doing half of the company’s work then presumably company CEOs are furiously lobbying remuneration committees to be paid less so further improving productivity. Hmm. Thought not.
I read the NYT article with interest and am struck with the parallels with machine translation and post-editing. All those "AI audit"-type jobs - if it follows the languages profession, I would put good money on those practitioners being paid at half the rate for AI auditing as they are today for doing the work the AI is expected to do, even though the responsibility is the same or greater.
Post-editing a machine translation requires the same skills as professional human translation plus in addition post-editing skills, an even greater eye for detail and concentration (especially because of the plausibility), and often takes the same amount of time as a human translation from scratch - and sometimes a human translation is actually quicker. Yet rates for post-editing are 50-60% of translation rates if you're lucky. And that's before you even consider the sameness of AI-slop - every contract, every report, every bit of code, all the same...
The languages profession is losing skilled people because they can earn more at Lidl (cf the evidence to the HoL committee on public sector interpreting). If the same practice happens across the "AI-audit" field, who is going to be willing to do the jobs with accountability?
One wonders if AI is doing half of the company’s work then presumably company CEOs are furiously lobbying remuneration committees to be paid less so further improving productivity. Hmm. Thought not.