Imagine you have a box of magnets, but instead of just north-south attraction, each magnet has a tunable dial that controls how strongly it pulls, which direction it pulls, and how far the pull reaches. Now imagine cooling those magnets until they barely move, pinning them in place with light, and watching what happens when you let them interact. That is the ultracold polar molecule program in one paragraph — and this review by Gregory and Cornish is the manual for that box. The committed claim here is pedagogical, not experimental: this paper consolidates the conceptual toolkit needed to understand why polar molecules are a serious quantum platform, covering the chain from molecule creation (Feshbach association plus STIRAP to the rovibrational ground state) through to controlled dipole-dipole interactions in optical tweezers and lattices. The 26-page, 20-figure treatment walks the reader from rotational structure and DC Stark shifts through microwave dressing and shielding, ending at quantum simulation of spin models and many-body physics. It is a tutorial review published in Contemporary Physics, not a new result — but its value lies in making the entire conceptual stack accessible in one place. The ladder context matters. This field has hit several genuine milestones in the last five years: the JILA group achieved quantum degeneracy of KRb molecules in 2024, while tweezer-based experiments at Durham and Harvard demonstrated two-qubit entangling gates between individual molecules. Gregory and Cornish (Durham) are among the groups who pioneered CaF and RbCs work in tweezers. The review consolidates these advances but does not itself present new data. Architecturally, the platform sits in the AMO (atomic, molecular, and optical) physics family, specifically the ultracold-molecules-in-optical-potentials branch. The key physical resource is the permanent electric dipole moment (~1 Debye for bialkali molecules), which gives 1/r³ interactions tunable by external DC and microwave fields. This is distinct from neutral-atom platforms (which rely on shorter-range van der Waals or Rydberg interactions) and trapped-ion platforms (which use Coulomb interactions). The anisotropy and tunability of the dipole-dipole interaction is the structural advantage; the difficulty of creating and cooling molecules is the structural cost. Integrity is straightforward for a review: the paper synthesizes published, peer-reviewed experimental results from multiple groups (JILA, Harvard, Durham, Imperial, Innsbruck). It does not present new benchmarks. The risk in a review is selective citation or overstatement of maturity — this one is reasonably balanced, acknowledging that bulk-gas cooling to degeneracy is recent and that error rates in molecular gates still lag behind neutral atoms and ions. The milestone picture is the most useful thing to extract. Molecular two-qubit gate fidelities are currently around 60% (Holland et al. 2023). Neutral atom gates have hit >99.5%. The threshold for useful quantum simulation with molecules — where the long-range, anisotropic interaction genuinely enables physics you can't simulate classically — requires scaling to ~50-100 molecules in a lattice or tweezer array with gate fidelities above ~95%. We are probably 3-7 years from that, depending heavily on cooling and coherence improvements. The obvious experiment not in this review: a head-to-head comparison of molecular versus Rydberg-atom quantum simulators on the same many-body Hamiltonian, benchmarked against classical DMRG or tensor-network methods. The reason it is absent is that the molecular platform is not yet at the scale where such a comparison is meaningful — the field is still building the machine, not racing it. That is the honest state of play, and to the authors' credit, they do not oversell it.
