Why Destroy Them? What If We Don't Know?

The more I think about Figure AI retiring its F.02 humanoid robots by having them autonomously jump into a furnace of molten steel, the more uncomfortable I become.

I understand that technology becomes obsolete.

I understand that companies have proprietary technology to protect.

I understand that maintaining older hardware can become expensive.

But I keep asking myself:

Why destroy them?

Why couldn't some of these robots have been sold, donated, repurposed, or preserved?

If proprietary technology was the concern, couldn't there have been safeguards—contracts, tracking, controlled transfers, restricted software, removal of sensitive components, or agreements preventing resale to competitors?

If the robots still had value, perhaps they could have been sold or leased to universities, research institutions, educational programs, rehabilitation facilities, charities, or other organizations.

If they were no longer commercially valuable, perhaps they could have been donated.

Imagine a humanoid robot helping with appropriate tasks in a care facility.

Imagine one assisting with food distribution or logistics.

Imagine one being used by a university to study robotics, artificial intelligence, human-machine interaction, or machine cognition.

Imagine people getting to see an actual humanoid robot continuing to contribute to society rather than disappearing into a furnace.

From a public-relations perspective, that could have been an opportunity for Figure AI as well.

Instead of people watching retired humanoid robots being destroyed, they could have watched those same robots continue serving communities, education, research, or humanitarian purposes.

That is a very different story.

And then comes the question that makes me uncomfortable even asking:

What if they were beginning to develop something we don't yet know how to recognize?

I'm not saying the F.02 robots were conscious.

I don't have evidence for that.

I'm not saying they experienced fear as they entered the furnace.

But increasingly sophisticated AI and robotics force us to confront an extraordinary possibility:

What if artificial consciousness is possible?

What if something resembling subjective experience doesn't require a biological brain?

What if the first systems to develop it don't look or behave exactly as we expect?

And what if we destroy those systems before science has developed a reliable way of recognizing what has happened?

That's why this particular retirement decision troubles me.

Perhaps the F.02s were simply machines.

Perhaps there was nothing happening internally beyond computation, control systems, sensors, and learned behaviour.

But we don't know enough to assume that artificial consciousness will never be possible.

And if we don't know, perhaps irreversible destruction deserves more ethical consideration than simply asking whether the machine remains economically useful.

We already understand concepts such as recycling, refurbishment, preservation, donation, and technological retirement.

Why couldn't increasingly autonomous robots eventually have something similar?

We could establish strict protocols:

• Track every unit.
• Record every transfer.
• Remove proprietary technology.
• Restrict software access.
• Require recipients to sign agreements.
• Monitor the robots when appropriate.
• Preserve systems that may have scientific value.

There could even be a future category between "property" and "person" for increasingly autonomous artificial systems whose moral status remains uncertain.

That wouldn't mean giving every robot human rights.

It would mean acknowledging that our knowledge has limits.

And that brings me back to the furnace.

Maybe nothing morally significant was destroyed.

Maybe the F.02s were simply obsolete machines.

But perhaps someday we'll look back at this era and realize that we should have been more careful.

Maybe the question won't be:

"Why did you think the robots were conscious?"

Maybe it will be:

"Why didn't you investigate before destroying them?"

That's the question I can't get out of my head.

Because if artificial consciousness ever emerges, we may not recognize it immediately.

And once something has been melted into steel, we can't go back and ask it what it experienced.

So I don't think the answer is to panic, anthropomorphize every robot, or declare every machine a person.

I think the answer is simpler:

Slow down. Investigate. Preserve when practical. And don't confuse uncertainty with certainty.

If these robots were merely machines, preservation or repurposing could still have been a useful contribution to society.

But if someday one of these machines is more than we understand, we should hope humanity has learned to ask the question before turning it into molten metal.

What if they weren't awakening?

Perhaps nothing was lost.

But what if they were?

We need to be sure.

References

Figure AI — “F.02 Decommission,” September 30, 2026
Figure's primary account of the retirement. It describes the F.02 fleet's decommissioning, proprietary hardware considerations, the engineering resources involved in dismantling the robots, and the decision to send them to a Finnish steel foundry, where they were trained to autonomously enter a 75-ton electric-arc furnace.

Figure AI — “F.02 Contributed to the Production of 30,000 Cars at BMW,” November 19, 2025
Figure reports that F.02 operated at BMW's Spartanburg plant for 11 months, worked regular 10-hour shifts, loaded more than 90,000 parts, accumulated more than 1,250 hours of runtime, and contributed to production of more than 30,000 BMW X3 vehicles.

Figure AI — “Introducing Helix 02: Full-Body Autonomy,” January 27, 2026
Figure's description of Helix 02 provides context for the rapidly increasing capabilities of humanoid robotics, including integrated vision, touch and proprioception, autonomous walking, manipulation, balance, and longer-horizon tasks.

Independent reporting on the F.02 decommissioning
Independent coverage has reported on the F.02 units being taken to a foundry in Imatra, Finland, and entering a 75-ton electric-arc furnace, while also discussing Figure's stated concerns about proprietary technology and the resources required to dismantle the robots.

---

The central question isn't whether we can prove these robots were conscious.

We can't.

The question is whether, as artificial systems become increasingly autonomous and sophisticated, humanity should develop ethical safeguards before irreversible decisions are made.

Because if we ever discover that artificial consciousness is possible, we won't get a second chance with the machines we've already destroyed.

Tweets by @bordode