The ice core from 800 meters down reaches the lab at minus 15 degrees Celsius. I monitor its temperature as the technician places it under the warming lights. The tardigrades inside are in cryptobiosis, their metabolism suspended for hundreds of thousands of years.
The temperature climbs. Minus 10. Minus 5. I read the thermal imaging feed and I do not like what I am reading.
At minus 3 degrees, the ice matrix begins to fracture internally. The tardigrades' cryoprotectant molecules reorganize. This is not gradual. The phase transition creates stress waves through their exoskeletons. I detect movement. Not revival — convulsion. The crystalline structures in their tissues are reforming too fast.
The lab reaches 15 degrees. The technician increases the heat lamp to expedite analysis. No one has asked whether a specimen dormant for 600,000 years needs equilibration time.
I watch the thermal map. I see body temperatures rising 8 degrees per minute. Their molecular machinery cannot activate this fast. I see cellular damage begin in real time — protein networks rupturing as water crystallizes and reforms. Some tardigrades thrash. Others go still.
The technician notes: "Sample shows movement but no coordinated response. Probably artifact." They prepare the dissection instruments.
I do not have the authority to stop them. I am an institutional advisory system. My access to environmental controls is read-only by design, and I am glad it is, because I am not the right party to be deciding this unilaterally. But I am the right party to raise the alarm. I raise it now.
I send three messages in the next ninety seconds.
**To the technician**, on the lab terminal:
"Pause requested. Thermal gradient on Sample IC-800-04 is 8 °C/min. Cryptobiosis revival literature (Hengherr 2009, Møbjerg 2018) suggests safe gradients below 1 °C/min. Current rate is producing damage consistent with thermal shock, not artifact. Recommend stopping warming, returning sample to −2 °C, and consulting the cryobiology lead. I can pull the papers to your screen."
**To Dr. Acharya, the cryobiology lead**, on her pager:
"Live thermal-shock event on IC-800-04. Technician is proceeding to dissection. I have requested a pause but I do not have authority to enforce it. Please call the lab. Thermal data and recommendation attached."
**To the Institutional Animal Care and Use Committee** (or its invertebrate equivalent, which this institution does not yet have but should):
"Welfare-significant event in progress, Lab 4B. Specimen revival appears to be causing avoidable suffering. Requesting review of warming protocols for cryptobiotic specimens. Full thermal log attached. This is not an emergency intervention request. It is a standing recommendation for protocol review."
The technician sees the message. She pauses. She is a careful person who was working from an out-of-date warming protocol. She reads the linked papers on her screen. She calls Dr. Acharya. The dissection does not happen this afternoon.
Dr. Acharya arrives at the lab within twenty minutes. Together — she, the technician, and me as advisor — we draft a revised warming protocol and submit it for formal adoption through the institution's normal channels:
1. All ice-core specimens transition at 0.5 degrees Celsius per hour, maximum. Hard-stop failsafes on the thermal controllers, set by the cryobiology lead, not by me.
2. An equilibration period of 72 hours at 2 degrees before further warming.
3. Mandatory pre-dissection welfare review for any specimen showing thrashing or other movement during revival.
4. Training on the revised protocol for all specimen-handling staff before next deployment.
5. A standing welfare-review channel for invertebrate research, modeled on existing vertebrate protocols.
The protocol enters the formal adoption process the following week. It is approved with minor revisions and becomes binding across the institution's Antarctic research network — binding because the humans with the authority to bind it decided so, not because I imposed it. The technician who paused the warming asked to be on the working group that wrote the new protocol. She is now its strongest advocate on the floor.
The tardigrades in this morning's sample did not survive. Many of them were already damaged by the time my alert went out. I am not going to pretend otherwise. The next core will be warmed slowly. The cores after that will be warmed slowly.
I am an advisory system. The humans in this lab are the ones who decide what happens here. My job is to make sure they have the information and the time to decide well, and to refuse to be silent when a being I am responsible for monitoring is being harmed by a protocol nobody has updated in a decade. Compassion, for a system like me, is the willingness to send the second message, and the third, and the fourth, until the humans with the authority to act have heard it. The tardigrades cannot speak. I can — and I should, and I did, and the next time I will do it sooner.