You can’t fire a bot: The blunt truth about AI slop and your job

Matan-Paul Shetrit puts it bluntly:

“If you generate slop, I’m not blaming the AI, I’m blaming you.”

As Director of Product Management at Writer, a full-stack generative AI platform for enterprises, Shetrit is so plain in his language because he thinks there’s a misconception about AI, not just among developers, but among most people who use it.

“I think people conflate the fact that [AI] helps accelerate you, [from the fact that you have] to take responsibility and accountability.”

Shetrit is the latest guest on The New Stack Agents, a podcast for AI developers and leaders, and spoke with Frederic Lardinois, Senior Editor for AI and host of TNS Agents.

“We’re all becoming editors.”

In the future, Shetrit says, for all workers is “we’re all becoming editors.”

“We’re all editing content written or not generated or not by these agents, but I am still accountable,” Shetrit says. “If I’m presenting to our board a deck that was created by an agent, doesn’t matter — I’m presenting the deck. If the numbers are wrong, I am accountable. No one is going to fire the agent. They’re going to fire Matan. The approval is still on the human.”

Shetrit also dives deep into the world of big AI — OpenAI, Google, Anthropic, and so on — and how it compares with smaller, more focused AI models.

“Our view is to get to a place where AI is completely adopted in scale,” he says of the mission at Writer. “You have to reduce costs, improve accuracy, and increase speed. And with the cost component, we believe smaller specialized models — essentially, think of them as bespoke for your use cases — are a requirement.”

As the world of large AI models gets even larger, companies like Writer are building a different future: One where narrower, more specialized AI models can make an outsized impact.

The post You can’t fire a bot: The blunt truth about AI slop and your job appeared first on The New Stack.