Artificial intelligence is no longer limited to chatbots, image generation and writing assistance. Scientists are increasingly exploring how AI can help solve complex problems in medicine, chemistry, physics and materials science. The next major technological breakthrough may not come from a smartphone or a computer, but from a laboratory powered by artificial intelligence.

At the DBS Asia Future Forward Conference in Singapore on October 12, 2026, TikTok CEO Shou Zi Chew highlighted the possibility that AI could accelerate major scientific discoveries over the next two to three years. He pointed to fields such as biotechnology, chemistry, physics and materials science as areas with significant potential. <Cite refs={["turn601782news4"]}/>

How Is AI Changing Scientific Research?

Traditional scientific research can require years of experiments, extensive calculations and repeated testing. Researchers often need to evaluate thousands of possible compounds or materials before finding a promising candidate.

AI can help narrow these possibilities by analysing large datasets, identifying patterns and predicting which experiments may be worth pursuing.

For example, researchers can use machine-learning models to examine molecular structures and estimate how different compounds might behave. Scientists can then use these predictions to decide which candidates deserve further laboratory testing.

AI does not replace the need for experiments, but it can help researchers focus their time and resources on the most promising options.

Could AI Help Discover New Medicines?

Drug discovery is one of the most promising applications of AI. Developing a medicine traditionally involves identifying a biological target, finding suitable compounds, testing their effects and conducting extensive clinical trials.

AI tools can help researchers analyse biological data, predict protein structures, identify potential drug candidates and estimate how molecules might interact with particular targets.

This could help speed up some early stages of drug development and reduce the number of unsuccessful candidates entering expensive testing processes.

However, discovering a promising molecule is only the beginning. A potential treatment must still undergo laboratory research, safety assessments and clinical trials before it can be approved for use. AI-generated predictions alone cannot establish that a medicine is safe or effective.

AI Is Also Transforming Materials Science

Beyond healthcare, artificial intelligence is being used to investigate materials that could improve batteries, semiconductors, solar panels and industrial manufacturing.

Researchers can use AI models to predict material properties and identify combinations that may offer better performance.

For instance, battery research involves balancing energy capacity, charging speed, durability, cost and safety. AI could help scientists evaluate a much larger range of possible material combinations than traditional methods allow.

If these predictions translate into successful experiments, the technology could eventually contribute to more efficient energy storage and improved electronic devices.

What About Physics and Other Scientific Fields?

AI is also being explored for analysing complex physical systems, processing experimental data and assisting with scientific simulations.

In astronomy, machine-learning systems can help identify patterns in large volumes of observational data. In physics, AI can assist with modelling complicated systems and detecting relationships that may be difficult to identify through conventional analysis.

The greatest benefit may come from combining AI's ability to process information at scale with scientists' understanding of how to design experiments and interpret results.

What Are the Risks of AI-Driven Scientific Discovery?

Despite its potential, AI-based research faces several challenges.

Accuracy: AI models can produce incorrect predictions, especially when working with incomplete or biased data.

Verification: Scientific claims must be tested through experiments and independent review.

Safety: Powerful AI systems could potentially be misused to support harmful research, making appropriate safeguards essential.

Access: Advanced computing resources and high-quality datasets remain expensive, which could concentrate scientific advantages among a limited number of organisations.

Researchers will need strong validation procedures, responsible oversight and transparent reporting to ensure that AI-assisted discoveries are trustworthy.

Will AI Replace Scientists?

AI is more likely to change how scientists work than eliminate the need for scientific expertise. Researchers will still need to define problems, develop hypotheses, design experiments, assess evidence and decide whether a discovery is meaningful.

The emerging model combines human judgement with AI's ability to analyse information and explore possibilities at scale.

If the technology continues to improve, AI could become an important research partner across laboratories, universities and pharmaceutical companies.

The biggest breakthrough may not be a single invention. It could be a fundamental change in how quickly humanity can investigate difficult questions, test new ideas and turn scientific knowledge into practical solutions.