In 2013, researchers at Google unleashed a descendant of it onto the text of the whole World Wide Web. Google’s algorithm turned each word into a “vector,” or point, in high-dimensional space. The vectors generated by the researchers’ program, word2vec, are eerily accurate: if you take the vector for “king” and subtract the vector for “man,” then add the vector for “woman,” the closest nearby vector is “queen.” Word vectors became the basis of a much improved Google Translate, and enabled the auto-completion of sentences in Gmail. Other companies, including Apple and Amazon, built similar systems. Eventually, researchers realized that the “vectorization” made popular by L.S.A. and word2vec could be used to map all sorts of things. Today’s facial-recognition systems have dimensions that represent the length of the nose and the curl of the lips, and faces are described using a string of coördinates in “face space.” Chess A.I.s use a similar trick to “vectorize” positions on the board. The technique has become so central to the field of artificial intelligence that, in 2017, a new, hundred-and-thirty-five-million-dollar A.I. research center in Toronto was named the Vector Institute. Matthew Botvinick, a professor at Princeton whose lab was across the hall from Norman’s, and who is now the head of neuroscience at DeepMind, Alphabet’s A.I. subsidiary, told me that distilling relevant similarities and differences into vectors was “the secret sauce underlying all of these A.I. advances.”
Subsequent sections of the article describe how machine learning has been brought to brain imaging with voxels of neural activity serving as dimensions in a kind of thought space.
...today’s thought-decoding researchers mostly look for specific thoughts that have been defined in advance. But a “general-purpose thought decoder,” Norman told me, is the next logical step for the research. Such a device could speak aloud a person’s thoughts, even if those thoughts have never been observed in an fMRI machine. In 2018, Botvinick, Norman’s hall mate, co-wrote a paper in the journal Nature Communications titled “Toward a Universal Decoder of Linguistic Meaning from Brain Activation.” Botvinick’s team had built a primitive form of what Norman described: a system that could decode novel sentences that subjects read silently to themselves. The system learned which brain patterns were evoked by certain words, and used that knowledge to guess which words were implied by the new patterns it encountered.