<i>unsophisticated linguistic algorithms + large amounts of data >= sophisticated linguistic algorithms + only a small amount of data.</i><p>Or<p>What you think is sophisticated, isn't.<p>How would you demonstrate which it is?
Large companies have known for a long time that your dataset is more important than your algorithms.<p>Data is facts and algorithms are basically opinions of the developer. A smarter developer may have more correct opinions on how data is related, but it will always be skewed by the limited perspective of the developer.
slightly off-topic, but relevant for the future of this technology.<p>I love Sci-Fi novels such as Peter Hamilton's "Reality Dysfunction" where an AI does the function of all local and central government in a habitat for no personal gain and every single person is connected to it real time on a personal basis.<p>The day we can move from the slimy self preserving low-life which seems to inhabit this space now will be a better day for everyone.<p>I wonder how close we are?