Big Tech wants more free stuff in AI copyright fight
Fetch another superyacht for the tech bros. As AI comes after copyright, Jack’s advice to wannabe creators — “embrace poverty, it’s easier that way” — is even more cogent.
Tech entrepreneurs make obscene profits. Lawyers have picnics; sometimes smorgasbords.
Politicians dine at troughs, virtually every moment of their lives subsidised by taxpayers. Pickings are a good deal slimmer at the other end of the table where creative people, amorphously known as content providers, are fed scraps.
It has always been thus but now in an interim report from the Productivity Commission under consideration by the Albanese government, creative people have been asked to forgo the fruits of their labour, in order to create a low or no regulatory environment around AI.
NewsCorp Lighthouse research, conducted by the Growth Intelligence Centre, found three out of four Australians believe tech companies should be prohibited from using original Australian content to train AI without permission or payment.
The survey, which fielded responses from 1276 people across the country, found 83 per cent of respondents believed the government should move to “protect Australian creators, ideas, stories, music and art”.
More than three quarters of respondents believed the government should maintain legal protections that protect “creative industries and content being used without compensation by foreign tech giants”.
And 77 per cent believed Australia should “stand firm if foreign companies or governments try to pressure us into giving up creative content for free or risk trade restrictions/tariffs” and 74 per cent agree big tech companies “should not be allowed to use original content without permission or payment”.
We might have guessed these responses to within a few points. Australians understand fairness and become passionate when they see their fellow Australians receiving the rough end of a pineapple.
This affects artists, authors, chroniclers — indeed anyone who has overcome the tyranny of a blank page or canvas or anyone who has rendered sound from silence.
Anyone who has sought to inform, report, entertain or amuse. Anyone who at the end of the day has put a ‘c’ in parenthesis alongside the year their work meets the public eye.
We all make choices and mine was to write.
Having done it for a long time, I am often asked for advice from aspiring scribblers.
“Embrace poverty. It will be easier that way.” Writers in Australia on average earn $18,000 a year, where the nation’s minimum wage for a 38-hour week annualised comes in at just under $50,000.
When a writer receives a decent advance on publication, and I have had a few, they do so knowing it is merely the provision of some cash against sales and royalties. I’ve done the maths after the completion of a manuscript.
Hundreds of hours of writing thousands of words. Dividing it by hours worked can be soul destroying. It would be more profitable to pop a sewing machine in the garage and take in piece work for the lower ends of the textile industry.
I have worked on both sides of the ledger as an author and a publisher.
Publishers absorb great risk. There is the inevitable prospect of commercial failure, or that a published work might trigger Australia’s absurd defamation laws.
Risk managing the exercise assumes great cost in legal advice and production.
In 2023, it became known that Meta, the owners of Facebook and Instagram had helped itself to the works of writers through LibGen, a shadowy Russian-based database of pirated works.
The purloined material was used to train Meta’s latest AI iteration, Llama.
In court filings it was shown that Meta had determined that it would take the works and let the intellectual property chips fall where they may. When questions were raised internally on how the project might proceed given copyright issues, comments made by ‘MZ’ pushed for the green light. Hurry up and break things.
Llama like almost all AI, open sourced or not, is designed to be monetised by tech companies. Fetch another superyacht for Zuckerberg. Throw another Picasso on the barbie.
Earlier this year, The Atlantic published a searchable database of authors’ works that Meta had plundered.
Two of my books had been shovelled into the Meta AI database years before without notice to the publisher or the author.
Is it theft? It sure feels like it. Collectively, it was the heist of the century in terms of intellectual property theft.
Millions of books, (apparently books are more sought after to ‘train’ AI than news articles) and academic papers had been yoinked.
Legal action was taken by a number of authors, including American comedian, Sarah Silverman. That action failed essentially on procedural grounds.
A ruling in June that Meta offered only more mystery around issues of fair use under US copyright law.
“This ruling does not stand for the proposition that Meta’s use of copyrighted materials to train its language models is lawful,” a US federal court judge determined. “It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one.”
In any event, individual content providers are not likely to engage in legal action against tech companies with billionaire owners in the US or here.
The notion of a regulatory light touch across AI in Australia, where intellectual property rights are rent asunder would be an act of reverse osmosis, draining from battlers to the super rich.
The Productivity Commission report paints a picture of economic benefits while ignoring a catalogue of dire social consequences, mostly associated with how we all work and hope to work in future.
The approach seems to be, let’s make some money until the cyborgs kick the door down.
To join the conversation, please log in. Don't have an account? Register
Join the conversation, you are commenting as Logout