In every species except man, the benefits of intelligence are largely limited to a single individual or a small family clan. Without widespread cooperative behavior there can be no evolutionary pressure to push intelligence to our heights. What use is it for a gorilla to read or have complex language if it can't be used to cooperate with a large number of other individuals?
Unfortunately, both Darwinian genetics and game theory tell us that cooperative behavior is exceedingly unlikely to emerge. In fact, there exists no well supported theory that fits the Darwinian genetic model of evolution for how non-kin cooperation could have emerged in the human species, let alone one that's widely accepted or which is supported by evidence. All the existing theories are predicated on the argument that cooperation is beneficial therefore evolutionary would select for cooperation, but neither genes nor evolution have foresight.
On the other hand, it should be abundantly clear that the types of intelligence which benefit the individual or a family clan arise readily in nature. Tool use, teaching, even culture are clearly widespread; they're simply circumscribed by well known genetic counter pressures. Cheating strategies spontaneously emerge and effectively limit the establishment of cooperative behaviors to close kin and, to a lesser extent, non-kin that are genetically incentivized to play along (e.g. mating partners or prospective mating partners), but only to the extent that have the incentive to play along. The two species of mammals which exhibit the greatest degree of systematic social cooperation (i.e. the ones closest to humans in terms of eusociality, though still not remotely comparable) are bonobos and mole-rats. Predictably, those groups are predominately composed of cooperating genetic sisters, so in fact their behavior is more analogous to bees and ants than to humans in terms of how and why eusociality exists.
It follows that the emergence of non-kin cooperative behavior likely preceded our intelligence. This is the bottleneck--not thumbs or speech or whatever phenotype du jure, including complex intelligence. None of this stuff stands up to scrutiny. All the evidence and theory clearly points to cooperative social behavior, for which there's little reason to believe was predicated on some simple phenotype. If only we knew how we passed through that evolutionary bottleneck....