Image analysis is the most established clinical application, with approved products in routine use.
What the systems do
Flagging findings for clinician attention.
Which is a support role rather than a diagnostic one.
Regulatory approval
Medical device pathways requiring evidence.
Which several products have completed.
Generalisation problems
Performance dropping on data from different equipment or populations.
Which is a documented and serious issue.
Workflow effects
Benefit depending on how the tool is integrated.
Why deployment is slower than capability
Regulatory approval requires evidence, integration into clinical workflow requires redesign, liability questions require answers, and clinicians reasonably require demonstrated benefit.
Which together mean a technically capable system can take years to reach routine use.
That process is frequently described as institutional inertia and is more accurately described as the normal standard applied to anything that affects patient outcomes.
Distribution shift
Systems trained on one population underperforming on another.
Which is the most documented practical failure.
Automation bias
Clinicians deferring to system output.
Which studies have measured.
Where benefits are clearest
Screening volume and triage prioritisation.
A general note
This describes technology rather than offering any medical guidance.
Why the mechanics are worth understanding
These systems are described in language that suggests understanding, reasoning and knowledge, and the underlying operations are considerably more specific than those words imply. That gap is where most confusion, most disappointment and a substantial amount of misplaced trust originates.
Someone who knows that a language model predicts tokens, that it has no separate lookup step, that its context is a hard limit and that its confidence is unrelated to its accuracy will make better decisions about when to use one than someone working from the marketing description. Not because the technology is worse than claimed, but because knowing what a tool actually does is how anyone uses a tool well.
The pattern across all of this
Nearly every behaviour that people find surprising follows directly from a design choice that is documented and comprehensible. Invented citations follow from the training objective. Difficulty with spelling follows from tokenisation. Degradation in long conversations follows from context limits and attention behaviour. Cost structures follow from the fact that inference is not free.
None of these are mysteries or failures of implementation. They are consequences, and they are predictable once the mechanism is understood.
Where the reliable information is
Research papers, model documentation, technical blogs from the organisations building these systems, and independent evaluation work. Most of it is freely available and considerably more measured than either the promotional material or the criticism that circulates.
The field moves quickly enough that any general article, including this one, dates rapidly. Anything with consequences attached is worth checking against current documentation rather than against a description written at some point in the past.
What follows from all this in practice
Use these systems for work where the output can be checked, where fluency is genuinely useful, and where a draft that needs revision is more valuable than a blank page. Avoid relying on them for specific facts, citations, numbers or anything where a confident wrong answer is worse than no answer.
That is a narrower recommendation than the enthusiasm suggests and a much broader one than the dismissal allows. Both positions tend to be argued by people who have not looked closely at what the systems do, and the useful territory between them is large.
On the pace of change
Capabilities have moved quickly enough that specific claims about what models can and cannot do have a short shelf life. Several limitations described confidently a few years ago have been substantially reduced, and several that were expected to fall have not.
The underlying mechanisms have changed far less than the capabilities. Next token prediction, tokenisation, context limits and the absence of a verification step have all remained, which is why the same categories of failure keep recurring in new forms even as performance improves.
A note on evaluating claims
When a capability is announced, useful questions are: measured on what benchmark, compared against what, evaluated by whom, and reproducible by anyone outside the organisation making the claim. Those four questions dispose of a considerable amount of what circulates.
Independent replication in this field is under-resourced relative to the scale of the claims, which is a structural problem rather than a criticism of any particular organisation.
Further reading
Technical documentation from model providers, peer-reviewed research, and independent evaluation organisations are the three places where this material is covered accurately.
All three are freely accessible, and all three are considerably drier and more useful than the commentary built on top of them.
One last observation
The most useful thing anyone can know about these systems is what they are optimising for, because everything else follows from it. A model trained to produce plausible text will produce plausible text, reliably and regardless of whether the text is true.
That is not a criticism. It is a specification, and reading it as one makes the whole field considerably easier to think about.