"How far do you think machine learning will be able to progress, do you see development hitting a wall in the future?"
There might be another period of plateau, where progress slows, but I definitely don't see AI hitting a wall because 1. computers continue to get more powerful 2. quantum computing is coming 3. human ingenuity is a major factor the recent successes, which involve increasing algorithmic sophistication 4. machine learning means algorithms are kind of writing themselves now... This process is likely to accelerate, like all technology.
Do you think machine learning in it's current state is already able to replace simple jobs? If not how long do you think it will take?
Not just now, but automation has been replacing jobs since the beginning of the industrial revolution. This condition led to neo-luddism. There was a report out of Oxford University from I think 2013 (I will link when I find it) predicting something like 40-50% persistent, long-term unemployment due to impending capabilities of contemporary AI.
"How will machine learning affect jobs? Do you see machine learning and AI taking over jobs like programming in the future or do you think it will create more jobs than it destroys?"
Automation doesn't just replace jobs, but also can lead to new jobs. Instead of working the assembly line, the humans now maintain the robots. Because Neural Networks are a bit of a black box (we have no idea what they're "thinking") it's possible there could be a whole new class of human jobs evaluating Neural Network performance, and even overseeing them in some way as a safeguard. There are probably lists out there of the type of in danger or being replaced. (I'll link if I find any:)