Mock up of an experiment teaching ants how to pipette.

Teaching Ants How to Pipette

Tuesday, November 28, 2023

Pipetting takes approximately 1027.66 hours/year-person (unpublished) in an average biology lab, and a way to automate that process would save the United States approximately $1.2 billion per year. Surprisingly, despite the scope of the issue there is a large gap in knowledge.

Ants are extraordinary social animals that are capable of performing amazing feats. Scientists have looked into using ants as tools but there have been few breakthroughs in ant tool use within the past 3 years. This epidemic of pipetting provides the opportunity to push the ant tool field forward in an exciting new way. 

However, incremental steps must be taken before we teach ants how to use pipettes. This experiment is to set up the framework and tools that will be used to teach ants to pipette.

We used humans as a model organism, as they are similar to ants in many respects; they are social animals that form complex colonies typically headed by a singular leader. Considering mouse studies are applicable to humans only about 20% of the time one could reasonably expect that a study in humans would be applicable to ants about 80% of the time.


Designing the pipette teaching implement

We designed a teaching implement in the size and shape of a pipette and the ability to use suction force to aspirate or dispense fluid with the push or pull of the plunger through a piston mechanism. To make the implement more realistic, we added a volume display and a volume change mechanism, connecting it to the piston to allow for different amounts of fluid to be aspirated or dispensed. 

These implements were calibrated so that the amount dispensed was accurate to the volume display. To be clear, this is not a pipette, this is a teaching implement specifically designed and manufactured for this experiment, but henceforth the implement will be referred to as a pipette for the sake of clarity.

Because animals are food motivated, we smeared peanut butter on the dispensing button of the pipette to encourage humans to press the button. To reward the behavior of pressing the button, we aspirated peanut butter through the pipette so that when the human presses the button they will be rewarded with peanut butter, encouraging the behavior.

Humans are able to operate pipettes

We obtained a swarm of humans through ethical means and put them in a room with the pipette. The first set of test subjects escaped but we quickly learned to lock the door and make sure the room does not have windows. The swarm appears agitated at first but eventually learns to push the button on the pipette. This result validates the hypothesis that the experimental method is capable of teaching humans how to operate a pipette.

Humans do not need to be in swarms to operate a pipette

To push the limits of this new method, we experimented with various parameters. With regards to the motivation mechanism, we found the results repeated with a variety of food, chemicals, as well as with imaginary things like ‘money’ and ‘college credits’. This is a testament to the robustness of this method of pipette training. We also experimented with the size of the human swarm, as the humans displayed a 2 on the grimace scale. Shockingly, the time to operate the pipette decreased with a smaller swarm, the fastest time being with individual humans.


We were able to get individual humans isolated from their swarm to operate a pipette, so using similar methods a singular ant should be able to operate a peanut butter covered pipette about 80% of the time. In addition, we found humans localized to laboratories had a significantly (p < 0.05) lower time, indicating a spatial dimension and suggesting that capturing ants localized to laboratories could yield similar results.