A disturbing discussion is brewing in Silicon Valley over whether machines should be granted the power to decide who lives or dies when at war, or whether artificial intelligence would merely be sown into weapon systems.
The debate gained ground in late September after Brandon Tseng, co-founder of Shield AI, advanced the argument that the U.S. would never have fully autonomous weapons wherein AI algorithms would not be the final authority over lethal actions.
He said Congress and the public do not have such a notion.
Shield AI is a US-based American aerospace and arms technology company based in San Diego, California.
It develops AI-powered fighter pilots, drones, and military technology
Just a few days later, Palmer Luckey, co-founder of Anduril expressed a contrasting view during a talk at Pepperdine University.
Anduril Industries is a US-based American defense technology company that develops cutting-edge autonomous systems.
He indicated his readiness to consider the utilization of self-driving weapons and had reservations regarding the moralistic arguments against it.
Palmer Luckey argues that people appear to be expecting to sidestep paradoxes of landmines, which cannot distinguish between innocent civilians and military targets; thus, ethical concerns around AI in warfare should be approached pragmatically.
According to a later statement by Anduril representative, Palmer Luckey's words did not say "robots should kill should be decided by the robots themselves. What he really communicated was his concern that the malicious actors would leverage mischievous AI technologies.
Some of the most influential tech leaders in the past, such as Trae Stephens, Anduril, would argue for a model in which humans make all critical decisions regarding lethality.
He believes accountable parties should be involved in the decision-making process.
While Anduril's representative denied that conflict exists between Luckey's and Stephens' views, the core feeling is that no matter what, someone has to be responsible in the chain of responsibility for the decision-making process when lethal force might be used.
The position of the US government on this question is equally unclear. The military does not currently buy fully autonomous weapons.
Though some assert that the capability of some systems to operate independently is all but autonomy, this does not come close to systems identifying, obtaining and attacking their own targets without human input.
The US prohibits neither fully-autonomous arms development nor the sale of such outside its borders
The US imposed new guidelines on AI safety in military applications last year with several allies endorsing them.
These rules obligate the highest military officers to ratify any new autonomous weapons, but it is entirely on a voluntary basis, and these officers have been on record with statements indicating that the time to think about a mandatory ban on autonomous weapons has not yet come.
Actually, not too long ago, Joe Lonsdale, co-founder of Palantir and an investor in Anduril, commented in the same vein, commenting on his readiness to explore the concept of fully autonomous weapons.
He refused to comply with the two-tiered view on this issue and persisted that a better understanding on autonomy in weaponry is necessary.
Activists and human rights groups had for long attempted to achieve international bans on autonomous lethal weapons, but to no avail as the U.S. opposed efforts on such.
Yet, all this might depend on a long conflict in Ukraine, everything thrown upside down, with lots of data about the use of AI in combat and a test ground for defense technology companies.
The Ukrainian officials say they need more weapon system automation-their belief is that this will power their capabilities against Russian forces.
What they fear, and this cuts across US officials and leaders in the tech industry, is that countries like China and Russia will rush into the development of fully autonomous weapons, and the US is going to be forced to catch up on its laggings.
The said fear was underlined by a Russian diplomat who spoke during a UN debate on AI arms, positing that priorities when it comes to human control differ greatly between nations.
In reaction to this competitive landscape, the tech leaders like Lonsdale are raising awareness about making proactive education effort toward military and governmental officials for better understanding of the benefit of AI toward national security.