Robo-debt was casually cruel, but we can’t get woolly about welfare

I have wondered recently, as I’ve experimented with large language models, whether AI could do a better job of governing than elected politicians. Artificial intelligence is very good at saying things that sound good but mean nothing, so it would nail the media interview side of things.

But I’d be worried about contaminating the emerging technology. Since ChatGPT has had regular contact with us ordinary folk, Stanford and UC Berkley scientists have found that it’s getting dumber. Just imagine then how quickly contact with the motivated reasoning of the political class could render an AI imbecilic. As it did with robo-debt.

What do we mean by “universal compassion”? Is it fair to keep generations of the same families relying on welfare?Credit: Peter Riches

Even beyond AI income-averaging, there is something very robotic about the whole welfare debate. Opposing political camps have been having the welfare argument for so many years that they seem to have forgotten that it’s about individuals.

AI could be writing the arguments that are most often heard. The best form of welfare is a job. Raising the rate of Newstart would reduce poverty and inequality in Australia. Children don’t get cheaper when they reach the age of [insert here the level to which campaigners are trying to raise the single-parenting payment]. Dole-bludgers are taking taxpayers for a ride.

These slogans have become tribal lore.

In Australia, it took the casual cruelty of robo-debt to expose how much the Coalition had embraced a generalisation that cast people relying on welfare as bludgers and rorters. But the opposite idea, that everyone who gets welfare is just an economic dynamo in straitened circumstances, is either naive or neglectful. It ignores the very different and complex situations in which people find themselves needing a social safety net and how sticky that net can be for them and their children when they try to get out of it.

In parliament last week, newbie Liberal MP Keith Wolahan, who represents the Victorian electorate of Menzies, proved he can do better than robotically repeating slogans when he acknowledged that the Coalition got it colossally wrong with robo-debt. Indeed, he went further and explained why robo-debt was the opposite of what the Liberal Party should stand for.

Wolahan described a welfare system that would be consistent with Liberal values. Compassion and incentive to work sit at the heart of it. That’s pretty much bipartisan. But what constitutes an incentive to work? And what is compassion? How can we tell if our welfare system is working if we’re not debating those questions? Compassion is not necessarily just the opposite of robo-debt.

Just over 20 years ago, I lived in a very disadvantaged part of Berlin. (I was skint and the rent was cheap.) While I bought the meagre appointments for my apartment from the equivalent of the Salvos, people in the area received fridges and various furnishings from the government. “What I’m entitled to,” as they explained it to me, in line with the German government’s definition of what it takes to live with dignity.

The problem only became apparent when I got to know my neighbours better. I worked in a pretty menial job and even that was a subject of envy. I was, as they say, upwardly mobile. Many of the people my age wanted to work but hadn’t the faintest idea how to. They were victims of intergenerational welfare. The government had thought its job was merely to ensure they and their parents had enough to live on, but it had neglected to give them the skills to become self-sufficient.

The unkindness of measures introduced in the name of universal compassion lies in the pained eyes of a young person who is unequipped to change their situation. My neighbours weren’t failures; they had been failed.

I think of them by name wherever the common welfare slogans get trotted out. Mutual obligation has become shorthand for punitive measures. To me, however, it means that we have an obligation to do more than just throw money at the unemployed to take their problems elsewhere.

It’s now often forgotten, but that’s why prime minister Julia Gillard cut the single-parenting payment. It wasn’t about punishing single families but trying to give kids in those households the opportunity to see a parental figure in work. Perhaps the policy failed, perhaps it was ill-thought-through. But simply reversing it is not enough. The intention – to help kids see a path out of poverty – is an objective that must be preserved, even if we have to look for another way to achieve it.

Another politician who couldn’t be replaced by AI is Jacqui Lambie. The colourful senator doesn’t have a coherent philosophical values system like Wolahan, but she doggedly insists on seeing people and articulating their problems. For instance, as she told the ABC, “two or three per cent of the population will never work in their life” because they’re just not up to it. No amount of punitive welfare-docking measures will make them capable of holding down a job.

Providing these people with enough to live is not all it takes to be compassionate. A good society needs to see that the stickiness of welfare can trap their children. Instead, it needs to come up with new ways to help the kids. I’m a big fan of long-day school, which could combine sport and one-on-one tutoring for those who need it with decent meals. Children should not be abandoned to rely on parents who are unable to help themselves.

It’s past time for a discussion on what a genuinely compassionate welfare system should look like. Robo-debt has shown that the combination of ideology and AI make us particularly stupid. Arguably, the AI wasn’t even necessary. Ideology can produce its own special type of idiot.

Parnell Palme McGuinness is managing director strategy and policy at award-winning campaigns firm Agenda C. She has done work for the Liberal Party and the German Greens.