In order for AI to function, it needs human intelligence and serious manhours
“Grunt workers.” This is a term one contractor refers to himself and others that have worked to build artificial intelligence companies like ChatGPT in recent years.
“You can design all the neural networks you want, you can get all the researchers involved you want, but without labellers, you have no ChatGPT. You have nothing,” the contractor reveals in an NBC news report, which found that many AI labellers working for companies such as OpenAI are making $15 an hour.
AI models are dependent on clean and labelled datasets, a process which is labour-intensive and often carried out by data enrichment contractors. That’s according to Partnership on AI, a nonprofit coalition of companies including Google, Meta, Amazon, Microsoft and IBM, that researches best practices for artificial intelligence systems.
It is not lost on most Freelance Informer readers that the people building AI and profiting from it are also the ones backing the non-profit writing the rules of conduct. That said, the organisation is showing signs of independent thinking.
“Despite the foundational role played by these data enrichment professionals, a growing body of research reveals the precarious working conditions these essential, but largely unseen, workers face,” says Partnership on AI in a report.
In its report it said,
While data enrichment work is getting increasingly sophisticated and requires more specialised skills, many have pointed out how this work is often undefined, unrecognised, and underappreciated.
This can partly be attributed to the way AI is marketed as a technological advance that can be more efficient than humans. Acknowledging the high labor costs necessary to train AI models runs counter to this narrative. However, a lack of recognition and meaningful career advancement opportunities is problematic for workers
There are already signs of disenchantment and burnout for these workers. Two OpenAI contractors spoke to NBC News about their work training the system behind ChatGPT. To read more, click here.