Overview: Write a brief memo on first, the best argument for allowing an AI program
and robots to use QALY utilitarianism to determine how to ration a future novel
pandemic vaccine (similar to the Moderna or Pfizer vaccines for covid-19) in the USA.
Then second, make the best argument against such a QALY utilitarian AI rationing
system, claiming some other method of rationing is morally superior. Finally, argue which
side wins. Each of the 3 arguments will require a different ethical or metaethical theory.
Instructions: Write a brief memo on the best argument for and against rationing a
vaccine for a novel pandemic virus (similar to the 1918 Spanish flu, or SARS-Cov-2 in
2019-2021) in the USA using a QALY utilitarian AI and suitably programmed robots. For
this argument, imagine a new strain of avian influenza that is resistant to previously
developed vaccines, but is equivalent to the 2021 Delta variant of SARS-Cov-2 in
transmissibility and lethality.
For this memo, assume there is such a novel influenza virus spreading by fall 2028. Also
assume, probably extremely optimistically (Links to an external site.), that on January 1,
2029 a new vaccine for the novel virus strain will begin initial distribution, with only 20
million doses to specially chosen groups (Links to an external site.), usually assumed to
be rationed first to some health care workers (with even those varying by state (Links to
an external site.)), and with only ~5 million new doses available for Americans each week
(for the estimated total population of 330 million Americans, with almost all of them
susceptible due to limited immunity). Each person will require 2 doses, so the initial
supply only covers 10 million people. Also, the rush job on approval through a EUA will
also mean that we will still not be certain about the long-term effectiveness or safety of
the vaccine when it is first administered, or how well it will work to produce immunity to
the virus once it mutates.
So, even under this optimistic scenario, through the end of April 2029 there would be
only an estimated 100 million doses available (thereby vaccinating only 50 million
people); herd immunity is estimated to require at least 80% of Americans being immune
(through vaccination or otherwise), or at least 264 million Americans. The CDC has plans
on prioritization (Links to an external site.) (based on a moral principle of ‘save the most
lives’, with some caveats; see also lecture, and updated version with some tweaks
here (Links to an external site.)). But is the CDC correct?
Should instead a ‘QALY utilitarian AI’ robot calculate who should receive the vaccine and
in what order, then use properly programmed robot vaccinators to administer the
vaccines (so human vaccinators don’t become infected or vaccinate the wrong people)?
So, your memo assignment: who should get the vaccine, and in what order? What is the
best criteria for vaccine priority? Specifically, should maximizing expected QALYs form
the criteria for rationing?
Rubric: 4 points for a plausible argument (valid, with plausibly true premises) that a
‘QALY utilitarian AI’ with properly programmed robot vaccinators is the morally best way
to ration a future novel pandemic vaccine; maximizing expected QALYs should form the
official government policy for vaccine prioritization.
4 points for a plausible argument (valid, with plausibly true premises) that a ‘QALY
utilitarian AI’ with properly programmed robot vaccinators is NOT the morally best way
to ration a future novel pandemic vaccine, and it should NOT form the official
government policy;that is, maximizing expected QALYs should not form the official
government policy for vaccine prioritization. In this argument, you need to identify and
argue for some other method of rationing vaccines as morally superior (whether that be
the CDC’s method, or some other method).
2 points for a third plausible argument (valid, with plausibly true premises) as to which
wins: why argument 1>2, or vice-versa.
