AI in chemistry and materials science is most credible when it tackles a clearly defined problem, draws on high-quality data, and produces results that can be checked experimentally. Chemistry World Business editor Phillip Broadwith makes that case in his 18 August 2026 opinion column, while warning that vague promises and fear of missing out can make AI look more capable than it is.
When does an AI claim in chemistry deserve confidence?
Broadwith’s test is practical rather than universal: ask whether the problem is well defined, whether the data are good enough for that task, and whether scientists can test and refine the model’s output. These are evaluation questions drawn from his argument, not a validated scoring system or guarantee of success.
As an Amazon Associate I earn from qualifying purchases.
- Problem: Is the target specific enough to judge whether the model has helped?
- Data: Are the training data sufficiently high-quality and relevant to the problem?
- Experimental check: Can the prediction be tested in the lab, with results used to improve subsequent work?
- Human expertise: Do chemists help interpret outputs and recognize where a prediction does not make sense?
Broadwith’s concern is that some investment in AI is encouraged by competitive fear of missing out or attractive but nebulous cost-cutting promises. That is his interpretation of adoption, not a measured finding about the whole industry. A persuasive claim should explain what work the tool changes and how the result is verified, rather than relying on the label “AI.”
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
What examples does Broadwith consider more persuasive?
Protein structure prediction
Broadwith points to protein-folding and structure-prediction models recognized by the 2024 Nobel Prize in Chemistry. He highlights the foundations behind this work: large libraries of crystal-structure data and years of algorithm development. The example shows how sustained data and methods can support a defined scientific task; it is not evidence that AI can solve chemistry problems generally. The column gives no numerical performance statistic for these models.
#1 Best Overall
Predicted materials put to a physical test
He also cites CuspAI’s work with partners to make and test materials predicted by its models. The important feature in his argument is the route from prediction to physical testing. The column does not name those partners or quantify the results, so the example illustrates an experimental feedback approach rather than establishing how often such predictions succeed.
What does the column say about AI and drug discovery?
Broadwith suggests AI may reduce time spent making molecules that fail in preclinical development. He also says AI has not yet had much impact on clinical-trial success rates, which he identifies as the major bottleneck. These are his assessments in an opinion column, not conclusions supported there by a systematic dataset or trial-level analysis. The distinction matters: a tool may help screen or prioritize candidates earlier without resolving the challenges that determine whether a treatment succeeds in clinical trials.
Rank #2
Why is Broadwith’s position cautious rather than dismissive?
His argument is not that AI has no place in chemistry. It is that usefulness depends on fit: an appropriately bounded problem, relevant data, and an experimental process that can expose errors and support refinement. Human chemical understanding remains part of that process, rather than something a model’s output makes unnecessary.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsBroadwith’s conclusion is cautiously optimistic. He argues that carefully considered AI and machine learning, combined with human chemical understanding and experimental verification, could help address difficult problems across industries that use chemistry. The column offers a framework for judging claims and examples of the approach; it does not establish an industry-wide measure of impact.
Quick Recap
Best Value
Rank #4
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
