I Was Wrong(ish) About the Robots. At Least the Bees Are Happy.

Reflecting on a Master’s debate, genAI whiplash, and why the next “human” connection we need might be with something that can’t speak.

Flashback

Thirty months ago, I was deep in an elective for my M.S. in Public Interest Technology at Arizona State. One of the required online debates for that global communication class was to address the implications of communicating with robots. Our prompt was to respond affirmatively or negatively with an informed response to the prompt: “In a globalizing world, we should be prepared to communicate more with robots.”

This was two summers ago—before generative AI (genAI) exploded and before large language models (LLMs) became an everyday utility. At the time, I was convinced that the line between robots as tools versus companions was a hill worth dying on.

Here’s what I wrote in 2023:

In a globalizing world, should we be prepared to communicate more with robots? My answer is a resounding no. Technology has always been a term used to describe tools that help humans achieve tasks easier, cheaper, or faster. Robots are a technology and should remain so; they should not be an end goal unto themselves, but a means to achieve end goals. Let’s examine this from two angles: 1) from the perspective of people who make robots work, i.e., the programmers and 2) from the perspective of people who use robots, i.e., the general public.

First, I will review programming through operational and philosophical lenses. As we have seen, programming robots is extremely inefficient. Though Gates (2007) ultimately is a proponent of robots in every home, even he has to admit that “very little of the programming code used in one machine can be applied to another” (p. 60). He also reveals that “programmers have struggled to figure out how to efficiently orchestrate code running on many different servers at the same time” (p. 64). The inefficiency of programming manifests in robots that cannot store and act upon data concurrently, understand varying natural speech patterns, nor recognize differences in Black faces (Kantayya, 2020). Why do we need robots “to grasp objects of varying sizes, textures and fragility” (Gates, 2007, p. 63)? Why do we need them to think the way humans think? Arguably, programmers do not simply code robots without external pressure or interest. The general public has an important stake in this matter as well.

Gates (2007) writes that “[t]he popularity of robots in fiction indicates that people are receptive to the idea that these machines will one day walk among us as helpers and even as companions” (p. 62). On the surface, it seems he may be right—people do enjoy watching films such as the Star Wars series. However, are people actually receptive to the idea? It cannot be ignored that films like Star Wars are, indeed, fiction. Filmgoers step into cinemas to be humored, provoked, or otherwise entertained for a few hours; their goal is not to see the world as it actually is. A true test of the public’s penchant for robot helpers can be found in the Henn-na Hotel in Nagasaki, Japan. There, all tasks from check-in to food delivery to turning off one’s room lights are accomplished with (velociraptor, humanoid, and more conventional) robots. While many guests enjoy the novelty for some time (Motherboard, 2015), most reported that some features malfunctioned, they felt disconnected from real people, and that the robots were eerie. It all points to humanity’s desire to feel connected to one another, and we cannot establish true connection from robots, no matter how elaborate they may be.

Shall we dedicate menial tasks for robots to complete? Perhaps. Should we prepare to communicate with them? No. What many people forget in this technological future is that, while we have the power to build robots, we also have the power to not build them. Ultimately, we alone have the ability to reason. So let us decide not to communicate more with robots.

Word count: 522

References

Gates, B. (2007). A robot in every home. Scientific American, 296(1), 58–65. https://doi.org/10.1038/scientificamerican0107-58

Kantayya, S. (Director). (2020). Coded Bias [Film]. 7th Empire Media.

Motherboard. (2015, October 23). Inside the Japanese hotel staffed by robots [Video]. YouTube.

The Honeybee and the Megaphone

Re-reading it now, I’m dumbfounded by how quickly the inefficiencies I discussed have been virtually eliminated by the advent of genAI. We didn’t just decide to communicate more with robots—we decided to ask them to write our poetry, debug our code, and increasingly mediate our relationship with the natural world. The Henn-na Hotel age of unsettling hardware has been replaced by the uncanny valley of empathic software.

As I gradually transition into the expedition industry, I notice I’m standing at the intersection of this kind of technology and the environments it’s meant to monitor. If I was wrong about our desire to communicate with robots, maybe I was also wrong about who they should be talking to.

Let me introduce you to the meliponini—the stingless honeybees of the Peruvian Amazon. Last week, a ruling in the municipalities of Satipo and Nauta granted these bees legal rights. They are now the world’s first insects with the right to exist, flourish, and be legally represented. [source]

The law is a response to a deadly collision of ecological crises. Scientists like Dr. Rosa Vásquez Espinoza have studied how these bees—who pollinate around 80% of Amazonian flora—are being decimated by a “vicious cycle” of invisibility. Because they weren’t on official red lists of extinction threats, they couldn’t get research funding, and because they didn’t have funding, they didn’t have the data to get on the lists.

While they waited in this research purgatory, the bees were being decimated by a trifecta of threats. Invasive “Africanized killer bees” began violently seizing their territory while industrial pesticides were discovered in hives hidden deep in the remote jungle. As the population waned, so too did a vital resource: their honey, a natural pharmacy containing hundreds of antiviral and antioxidant molecules.

Turning a “No” into a “Maybe”

In my class at ASU, I argued that we shouldn’t communicate with robots because it diminishes human connection. But in the Peruvian Amazon case, there’s a communication gap that we haven’t been able to bridge. There’s a species that is voiceless in our legal systems, even though they are (quite literally) the glue holding the rainforest together.

Instead of using LLMs to mimic human companions, we could easily use them to act as translators for the voiceless. Groups like the Earth Species Project are already using AI to decode animal communication. What if we gave the bees a persona built on the chemical data Dr. Vásquez Espinoza collected? One that can interpret hive health into a legal memo that explains to a judge exactly how a certain pesticide infringes on their right to a “healthy habitat free from pollution”?

I stand by the idea that true connection belongs among humans over robots. But as an educator and technologist heading to the expedition field, I’m acknowledging that the robotic companions I once dismissed might be exactly what we need to really hear nature.