I would define sentience as having the ability to interact with the environment, show feelings, and react based on the interactions. Having some form of reaction would mean there are some motor skills involved, and having the ability to coordinate motor operations would require intelligence. Having the ability to hold information, learn from it over time and iterate over it.
Can an AI technology achieve consciousness? Let me take the bear case first
Because what we define as consciousness is very difficult to achieve, Neuroscience barely has any answers as to what makes something conscious so naturally if we don’t have a define an end target of what we want to build we will not be able to create it. Knowing what creates consciousness would give us a better direction of creating it.
Now for the bull case can an AI technology achieve consciousness?
Learning and reacting are not hard problems in and by themselves, we learn all our lives and develop our emotions over time. I don’t see why intelligent systems cannot mimic that to an extent, at which point their intelligence may not be mimicking our own intelligence but it is showing some form of intelligence. Having intelligence is a building block to exhibiting consciousness.
Intelligence will come in various shapes and forms, right now we see some of it. Sure an AI system will not learn to react from its inception, but they can learn this behavior over time just as many other beings do.