What is the value free ideal and should scientists strive to uphold it?

The value-free ideal states that non-epistemic values (ethical, political, etc. considerations) ought not to influence scientific reasoning. However, there seems to be a near consensus in philosophy rejecting this ideal, particularly on the grounds that it isn’t achievable (see The New Demarcation Problem).

Nevertheless, achievability does not deny desirability. This is similar to other ideals: the fact that it is virtually impossible not to lie at every conceivable instance does not mean that we should not strive to uphold the ideal ‘not to lie’. Moreover, it can be necessary or morally required to deviate from the ideal.

Rejecting the value-free ideal can lead to detrimental consequences, including the transformation of scientific inquiry into just another space for an ideological battle. This might prove true beyond pure science, e.g. as has been shown in overcompensations in large language models producing historically inaccurate images (see e.g. Black Nazis? A woman pope? That's just the start of Google's AI problem).

Therefore, even if not achievable, the value-free ideal should be retained in its original function as guidance for scientific research by reframing the ideal as a constraint in resorting to non-epistemic values and choosing among values.

For a detailed argument, see the essay that I wrote for the course 'Evidence and Policy' as part of the programme in philosophy and public policy at the London School of Economics and Political Science (LSE): https://philsci-archive.pitt.edu/id/eprint/24942.