My view is different from this, only to the extent that if a decision is taken, by the parents and doctors, that it is better that a baby should die, I believe it should be possible to carry out that decision, not only by withholding or withdrawing life-support - which can lead to the baby dying slowly from dehydration or from an infection - but also by taking active steps to end the baby's life swiftly and humanely.

There are a lot of weapons that we've developed which we've pulled back from - biological weapons, chemical weapons, etc. This may be the case with armed autonomous robotics, where we ultimately pull back from them.

...the nervous systems of other animals were not artificially constructed - as a robot might be artificially constructed - to mimic the pain behavior of humans. A capacity to feel pain obviously enhances a species' prospects of survival...it is surely unreasonable to suppose that nervous systems that are virtually identical physiologically, have a common origin and a common evolutionary function, and result in similar forms of behavior in similar circumstances should actually operate in an entirely different manner on the level of subjective feelings.

Grain that is used to feed animals that end up on our tables as turkeys and hams could have gone to feed starving people.

There could conceivably be circumstances in which an experiment on an animal stands to reduce suffering so much that it would be permissible to carry it out even if it involved harm to the animal... [even if] the animal were a human being.

In appropriate circumstances we are justified in using humans to achieve goals (or the goal of assisting animals).

Surely there will be some nonhuman animals whose lives, by any standards, are more valuable than the lives of some humans.