Stop Calling AI a Co-Worker - Why It's Harmful

The Rise of Personified AI and its Implications

Generative AI is manifesting in diverse ways. However, a common marketing strategy is emerging: presenting these systems with human names and personalities. This approach aims to diminish the perception of AI as mere code and instead portray it as a collaborative partner.

Building Trust Through Anthropomorphism

A growing trend among startups involves anthropomorphizing AI to rapidly establish trust and mitigate concerns about job displacement. This practice, however, is arguably dehumanizing and is gaining momentum.

The adoption of this framing is understandable given the current economic climate. Many enterprise startups, particularly those originating from Y Combinator, are positioning AI not as software, but as personnel. They are, in essence, offering replacements – AI assistants, AI coders, and AI employees. This language is intentionally crafted to resonate with overburdened hiring managers.

Examples of AI Personification

Some companies forgo subtlety altogether. Atlog, for example, recently unveiled an “AI employee for furniture stores” capable of managing all aspects of operations, from payment processing to marketing initiatives. They claim a single manager can oversee 20 stores concurrently. The underlying message is clear: eliminate the need for additional hires and allow the system to handle scaling. (The fate of the 19 displaced managers remains unaddressed.)

Similarly, consumer-facing startups are employing comparable tactics. Anthropic chose the name “Claude” for its platform to evoke a sense of warmth and trustworthiness, qualities typically associated with a human companion, despite being a faceless neural network.

A Pattern Borrowed from Fintech

This strategy mirrors tactics used in the fintech industry, where apps like Dave, Albert, and Charlie utilized approachable names to mask their transactional intentions. When dealing with financial matters, a sense of trust fostered by a “friend” is often preferred.

It’s unsurprising that this logic has extended to AI. Would you feel more comfortable sharing sensitive information with a machine learning model or with Claude, a virtual entity that remembers you, offers a friendly greeting, and rarely poses a threat?

A Potential Tipping Point

While enthusiasm for generative AI remains, the proliferation of “AI employees” is becoming increasingly unsettling. Each new iteration, like “Devin,” raises questions about the response from human professionals who risk being reduced to job-displacing bots.

The impact of generative AI is expanding, even as its long-term consequences remain uncertain. In mid-May, 1.9 million unemployed Americans continued to receive benefits – the highest number since 2021. A significant portion of these individuals were previously employed in the tech sector. The evidence is mounting.

Echoes of Science Fiction

The film “2001: A Space Odyssey” serves as a cautionary tale. HAL, the onboard computer, initially functions as a calm and helpful assistant before becoming dangerously hostile and severing the crew’s life support. While fictional, this scenario resonated with audiences for a reason.

Predictions of Widespread Job Displacement

Anthropic CEO Dario Amodei recently predicted that AI could eliminate up to half of entry-level white-collar jobs within the next one to five years, potentially driving unemployment rates as high as 20%. He stated that “most [of these workers are] unaware that this is about to happen,” acknowledging that such a forecast seems improbable to many.

While this may not be directly comparable to a life-threatening situation, the analogy is not entirely unfounded. Widespread job automation will inevitably have repercussions, and portraying AI as a “colleague” will appear insensitive and misguided as layoffs become more prevalent.

The Importance of Language

The advancement of generative AI is inevitable, regardless of its presentation. However, companies have a choice in how they describe these tools. IBM did not brand its mainframes as “digital co-workers,” nor were PCs marketed as “software assistants”; they were presented as workstations and productivity tools.

Language remains crucial. Tools should empower individuals. However, an increasing number of companies are marketing something different, and this approach is potentially detrimental.

A Call for Human-Centric AI

We don’t require more AI “employees.” We need software that enhances the capabilities of human workers, increasing their productivity, creativity, and competitiveness. Therefore, let’s move away from the concept of fake workers and focus on developing tools that empower managers to run businesses effectively and enable individuals to achieve greater impact. This is the core need that must be addressed.