Many researchers believe that the singularity – roughly, explosive growth in machine intelligence, resulting in the appearance of AIs that are vastly superior to us cognitively – will or might occur in the relatively near future. Understandably, many are worried by this prospect. In this paper, I argue that we have less to fear from the singularity than many people think. Depending on how we define 'singularity', it is either less likely to occur, or it will likely occur in a form that is not threatening to us. I argue that the capacity for cumulative culture is central to our success as a species, and AIs that rely on processing power alone, without the scaffolding of culture, are unlikely to outcompete us intellectually. There is, however, nothing to prevent AIs from having, and taking advantage of, a capacity for culture. AIs with such a capacity may have intellectual powers greater than ours currently are, but the collective deliberation that underlies the power of cumulative culture is more powerful when there is sufficient diversity among the deliberators. This fact will give AIs a reason to value our continued flourishing, so that we are able to contribute to valuable epistemic diversity.
|Number of pages||12|
|Journal||Science, religion and culture|
|Publication status||Published - 2016|