The laws of physics don't work in isolation. They always apply all at the same time¹.
Most of them don't describe causality, but relations that are known to be always satisfied². If you know the speed increased, you can conclude the pressure decreased, but if you know pressure decreased, you can similarly conclude the speed increased. Neither is the reason for the other, rather the properties of fluids are reason both always happen together.
And last but not least the laws of physics as we normally use them are not fundamental, but are various perspectives on the underlying properties of the universe and as such have a lot of overlap. Therefore you can usually arrive at the same conclusions in many different ways (provided you properly account for approximations involved).
In this case, the air following the surface of the wing, pressure dropping, the flow accelerating and the streamlines squeezing are all aspects of the one behaviour of fluids—for which there is no simpler way to quantify than calculating the full set of Navier–Stokes equations (there are some simpler models like Kutta–Joukowski theorem, but can't start from the wing shape, but need some experimentally-determined characteristic values of it).
On the low level, pressure is the total momentum imparted by the collisions per unit time and area. So obviously if lower pressure is seen, the collisions must be fewer or weaker. In practice they are both³. However, there is no simple explanation for it on the molecular level either. It is a result of the complex dynamics that averages out to the Navier–Stokes equation on the macroscopic level.
¹ School lessons often fail to teach this by discussing each in isolation and never synthesizing. However, lift is surprisingly complex phenomenon, which requires applying many of them to model.
² Well, at least not observed not to be satisfied, because positive proofs are not possible.
³ The enthalpy decreases, which means the particles move slower and that means they hit both less often, and with less force. The density also decreases a bit, so there is a bit fewer of them per unit volume too. However, this is all easiest to derive from the macroscopic thermodynamic laws. We do have microscopic explanations for them now, but trying to compute anything directly at the low level is futile due to the vast numbers of particles involved.