Sadly, I'm poorly read in the classics, so I'm not too familiar with Keynes' actual writings. However, I can give a superficial overview of what the modern econ literature says about the effect of automation on labor.

A cursory scan of Keyne's 'Economic Possibilities for our Grandchildren' suggests that he was concerned about technology effectively rendering labor obsolete and an end to subsistence scarcity.

Let's focus on the first part. Oversimplifying things a bit, suppose the labor market is made up of two groups of workers: low skill and high skill. In Keynes' time (and for the most part in modern times) automation substitutes some sort of non-human capital for low skill labor. Unsurprisingly, this depresses the wages of low skill labor, since the demand for low skill labor has dropped because of the availability of machines that do the same work.

But what happens to high skill workers? The demand for (and wages of) high skill workers
can increase. This happens because high skill work and low skill work are considered to be imperfect substitutes, or even complements. Let's say they are complements, because I think it makes the mechanism more intuitive. To explain what complements are, think of a doctor's office. The doctor is the high skill worker, and the receptionist is the low skill worker (relatively speaking). If there's no receptionist there working, the doctor probably not seeing many (if any) patients. The doctor needs administrative work to be done in order to do their own work. In this example, if the receptionist is replaced with the brand-new robo receptionist, it can actually makes the doctor more productive because all the extra admin work being done allows the doctor to do more work than before. Alternatively, it can increase the demand for high skill work if the robo receptionist needs a high skill worker to operate or maintain it. This type of technological change is called 'skill biased' because it favors a certain level of skill.

So, going back to Keynes' concern, automation might render certain forms of work obsolete, but not all, and can create new types of work for humans to do. There is a famous idea advanced by Tinbergen back in 1974 and 1975 that there is a 'race' between technology and the 'supply of [advanced] skills', that occurs with skill biased technological change. So while technology has advanced since Keynes' time, so has the supply of skilled labor, with the average level of education having risen considerably since then.

This is basically a poor attempt to put the models from Acemoglu and Autor (2011) into words. It's only tangentially related to my research area, so hopefully I didn't butcher it too badly. This is still an active research area, particularly with concerns over A.I. Interested readers might want to check out Acemoglu and Restrepo (2019) in the Journal of Economic Perspectives. For non-economists, this is a highly respected journal that is freely available where the articles have been written for non-technical readers, and Acemoglu is a luminary in econ, and probable future Nobel Memorial Prize winner.

The second part of Keynes' essay about the end to subsistence scarcity is something I definitely can't answer. It's just too multifaceted. However, there are certain research areas that are very likely related.

Inequality is a major one, while we've gotten much richer as a species, the allocation of these gains have obviously been highly skewed.

There's been quite a bit of work trying to understand why, as rich nations became rich, some countries such as the US saw average hours of work hold steady, while others in Europe saw them fall. In microecon theory this could be attributed to preferences between goods and leisure and the magnitude of substitution vs income effects.

There is also a stream of research on product variety that might explain part of why people still work long hours. Basically, people seem to like variety, and there have been a lot of new goods since Keynes' time. Perhaps a person could only work a few hours a day and live a comfortable, austere life. But if they work more, they might afford that new smart phone and gaming console, and the countless other new best thing that come out all the time now.

References:

Acemoglu, D., & Autor, D. (2011). Skills, tasks and technologies: Implications for employment and earnings. In Handbook of labor economics (Vol. 4, pp. 1043-1171). Elsevier.

Acemoglu, D., & Restrepo, P. (2019). Automation and new tasks: How technology displaces and reinstates labor. Journal of Economic Perspectives, 33(2), 3-30.