NLP seems easier to find the overlap of pay and good work. NLP on medical texts for instance is a very lucrative space that has enormous value to end users and

NLP seems easier to find the overlap of pay and good work. NLP on medical texts for instance is a very lucrative space that has enormous value to end users and society NLP for interpretation and foreign policy as well NLP for legal discovery NLP for education These are all spaces where you are ability to contribute meaningfully is very high, and the training time is surprisingly low for utilities. Theory it becomes very high, fundamentals it becomes very high training times In any sub industry you'll still have to avoid corporate chatbots or their equivalents