The Great Convergence: How AI Connects Every Frontier

Artificial Intelligence (AI) has set the world ablaze with its potential to enhance efficiency, reduce costs, and boost productivity.

While for many people AI translates to chatbots, thanks to the accessibility and popularity of generative AI tools like ChatGPT, the technology is far more expansive, with its benefits stretching into medicine, manufacturing, robotics, healthcare, education, climate science, finance, law, cybersecurity, and beyond.

By emulating human cognitive functions like problem-solving and decision-making, AI promises to transform the underlying systems of these industries, where a growing number of organizations are actively exploring AI’s capabilities.

A recent McKinsey survey revealed an increase in AI usage, not just in the technology sector, where it has already surpassed 90%, but in almost every industry.

Nine out of ten respondents said that their organizations are regularly using AI, though it is still in the experimentation phase. Despite being in the piloting stage, respondents reported cost and revenue benefits, with 64% saying that AI is enabling them to innovate.

This growing adoption shows that even at such an early stage, AI is becoming a key enabler of digital transformation.

In today’s globally interconnected and competitive landscape, AI is allowing businesses to harness the power of various digital technologies, such as big data, cloud computing, and the Internet of Things. In effect, it is acting as a converging technology, accelerating the development and integration of other technologies so that their combined impact is greater than the sum of their parts.

With that, let’s now take a look at some fascinating AI advancements across different sectors, with each one showing just how it is changing the world.

Below is a quick snapshot of three domains where AI’s convergence is already measurable.

Swipe to scroll →

AI Maps the Gut Microbiome to Human Health (and CAD Risk)

With the help of AI, scientists have now decoded the intricate ecosystem of gut bacteria and their chemical signals, allowing for the uncovering of hidden links between the bacteria and human health. What’s more is that the new advanced AI system is actually better in studies of cancer, obesity, and sleep disorders than conventional models, and showcases huge potential in customizing treatments on each person’s microbial makeup to transform personalized medicine.

AI’s capability in revealing hidden patterns is particularly remarkable, as in a study from the University of Waterloo¹, AI-powered analysis of routine blood tests detected patterns that can make life-saving predictions both affordable and accessible.

Coming back to gut bacteria, they play a key role in our health, influencing not just our digestion and disease prevention but also immunity and even our mood. New research has found that our gut microbiome² may also influence the development of coronary artery disease, which kills almost 20 million people every year.

The human gut is clearly fascinating, but it is also a complex ecosystem of trillions of microorganisms. Besides the sheer number of bacterial species present in our gut, their interactions with human chemistry make it challenging for scientists to understand their effects.

But in a groundbreaking step, researchers from the University of Tokyo turned to AI to tackle this issue.

They created an AI system to better understand which bacteria produce which metabolites, small molecules that act as chemical messengers and circulate through our body, influencing metabolism, immunity, and brain function, and how the relationship between bacteria and metabolites changes in different diseases.

“By accurately mapping these bacteria-chemical relationships, we could potentially develop personalized treatments. Imagine being able to grow a specific bacterium to produce beneficial human metabolites or designing targeted therapies that modify these metabolites to treat diseases.”

– Project Researcher Tung Dang from the Tsunoda lab in the Department of Biological Sciences

The model they developed is a Bayesian neural network called VBayesMM³, which addresses the challenge of identifying meaningful patterns from the complex interaction between trillions of bacteria and metabolites.

It uses a Bayesian approach to identify just which bacterial groups are influencing particular metabolites the most. Moreover, it measures uncertainty in its predictions to prevent any incorrect conclusions, which provides scientists with more accurate and trustworthy insights.

Using its variational Bayesian microbiome multiomics (VBayesMM) approach, the team has been able to identify key microbial species both rapidly and precisely, resulting in more accurate estimations. The implementation of variational inference, meanwhile, addressed computational bottlenecks, allowing for scalable analysis of massive datasets.

The team will next be working with more comprehensive chemical datasets to capture the complete range of bacterial products to overcome the problem of a drop in accuracy when metabolite data is more extensive than bacterial data.

“We also aim to make VBayesMM more robust when analyzing diverse patient populations, incorporating bacterial’ family tree’ relationships to make better predictions, and further reducing the computational time needed for analysis,” said Dang. “For clinical applications, the ultimate goal is identifying specific bacterial targets for treatments or dietary interventions that could actually help patients, moving from basic research toward practical medical applications.”

Investable Angle: Precision Medicine with Tempus AI (TEM +0.33%)

In this exciting and complex world of life sciences, Tempus AI stands out for offering AI-enabled precision medicine solutions for personalized patient care.

Tempus is a technology company advancing precision medicine as well as facilitating the discovery and development of optimal therapeutics. It has three product lines:

• Genomics: Provides next-generation sequencing (NGS) diagnostics, profiling, molecular genotyping, and other testing.
• Data: Involves structuring and de-identifying the data generated in its lab before commercialization.
• AI Applications: Provides diagnostics, implements new software as a medical device, and deploys clinical decision support tools.

This year, Tempus achieved a few key regulatory milestones, including receiving FDA clearance for the Tempus xR IVD device to support drug development through advanced RNA sequencing. As a result, Tempus partners can use its RNA assay to “more precisely identify just which patients are most likely to respond to specific therapies and to design more efficient clinical trials.”

Its updated AI-powered cardiac image analysis platform, Tempus Pixel, and AI software Tempus ECG-Low EF also obtained FDA 510(k) approvals, strengthening the company’s position in AI-driven diagnostics.

The $12.73 billion market cap company’s shares, meanwhile, are currently trading at $72.52, up almost 112% this year. Just last month, TEM shares surpassed the $100 mark.

As for its financial position, Tempus recently reported an 84.7% YoY increase in revenue in Q3 2025 to $334.2 million, while gross profit increased by 98.4% to $209.9 million. Net loss for the quarter was $80 million. It ended the quarter with $764.3 million of cash and marketable securities.

“Not only are we growing at an incredible rate, reaching positive adjusted EBITDA marks an important milestone and reflects the strength of our underlying business,” said Tempus founder and CEO Eric Lefkofsky. “One of the hardest things to do, and a sign of business model endurance, is being able to slow down the rate of reinvesting back into the business and still maintain growth, which is exactly what we achieved this quarter.”

Click here for a list of top biotech big data companies.

AI Forecasts Solar Storms Days Ahead—Protecting Power Grids & Satellites

An AI model has been developed to predict solar wind well in advance with greater accuracy than existing methods, helping protect power grids, satellites, and navigation systems from disruptive space events and strengthening the resilience of our critical infrastructure.

Solar wind is the constant stream of charged particles released by the Sun. It happens when the twisted magnetic fields of the Sun get contorted and stretched, leading them to snap before reconnecting, during which they release large amounts of energy.

When these particles pick up speed, they can disrupt Earth’s atmosphere. Not only can they interfere with power grids, but they can also drag satellites out of orbit, as a strong solar wind event did in 2022, when SpaceX lost up to 40 of its Starlink satellites.

Solar storms, meanwhile, are even more powerful events, where the sun blasts energy, particles, and magnetic fields into the galaxy. When directed towards Earth, it can create a major disturbance in its magnetic field, which is called a geomagnetic storm. This is what paints beautiful aurora borealis displays, but also causes power outages.

Lloyd’s latest systemic-risk scenario estimates that a severe solar storm could expose the global economy to about $2.4 trillion in losses over five years, with expected losses of roughly $17 billion today.

This shows an urgent need for better forecasting of these events. So, researchers at NYU Abu Dhabi (NYUAD) went on to do just that, with the help of AI.

They have built an AI model that can predict solar wind⁴ as many as four days before the event occurs, more accurately than current methods. The model has been trained on historical records of solar wind and ultraviolet images from NASA’s Solar Dynamics Observatory (SDO).

It’s by analyzing the images of the Sun to detect patterns linked to changes in solar wind that the NYUAD team was able to achieve a 45% improvement in their forecast accuracy compared to current operational models. Moreover, they achieved a 20% improvement over previous AI-based approaches.

“This is a major step forward in protecting the satellites, navigation systems, and power infrastructure that modern life depends on,” said the study’s lead author, Dattaraj Dhuri. “By combining advanced AI with solar observations, we can give early warnings that help safeguard critical technology on Earth and in space.”

Investable Angle: Space-Weather AI with IBM (IBM -1.9%)

The $293.24 billion market cap IBM is a global provider of hybrid cloud and AI services to help achieve digital transformation across data, applications, and environments in which they operate.

A couple of months ago, IBM released its open-source AI model, ‘Surya’, in collaboration with NASA, to better understand data collected from solar observations and predict how solar activity affects space technology and Earth. With Surya, the company is applying AI to space weather forecasting research and providing a tool to help protect telecommunications, power grids, and GPS navigation from the disruptions caused by the Sun’s changing nature.

As of writing, IBM shares are trading at $319, up 42.7% YTD. It has an EPS (TTM) of 8.07 and a P/E (TTM) of 38.87. IBM pays a dividend yield of 2.14%.

For its recent quarter, 3Q25, the company reported a 9% increase in revenue to $16.3 billion. Its GAAP gross profit margin was 57.3% and its non-GAAP operating profit was 58.7%. Net cash from operating activities, meanwhile, was $9.2 billion year to date, and $7.2 billion was reported in free cash flow.

“Clients globally continue to leverage our technology and domain expertise to drive productivity in their operations and deliver real business value with AI.”

– CEO Arvind Krishna

He also noted that IBM’s AI book of business exceeded $9.5 billion, up from $7.5 billion in the previous quarter.

Where AI Outperforms Experts: Medicine, Neuroscience & Education

Researchers are finding that AI consistently outperforms experts across various domains.

One such domain is medicine, where “glaucoma remains one of the most common causes of vision loss that can’t be repaired globally,” with screening being too expensive. But AI could be the solution to this, hopes Dr. Anthony Khawaja, a professor at the University College London Institute of Ophthalmology and the lead researcher of the new study, which reported that a trained AI program correctly identified patients with glaucoma about 90% of the time, compared with 81% for human graders.

For this study, both human experts and an AI program evaluated over 6,300 participants, of whom almost 700 had glaucoma in at least one eye.

Glaucoma is the result of damage to the optic nerve, most commonly due to pressure building inside the eye, which can lead to complete blindness. Human experts and the AI graded participants’ glaucoma risk based on vertical cup-disk ratio, a key measure of this disorder, which tracks changes in eye structure caused by pressure buildup.

According to the study results, only 11% of participants’ eyes were suspected to have glaucoma, matching the proportion expected in routine screening. The accuracy, researchers noted, could be further improved by including other risk factors, such as intraocular pressure.

In another study⁵, large language models (LLMs) predicted the outcomes of proposed neuroscience studies more accurately than human experts, highlighting AI’s potential to accelerate research.

Instead of focusing on LLMs’ question-answering capabilities, the study explored whether the models can synthesise knowledge to predict future outcomes.

So, they tested 15 different general-purpose LLMs and 171 human neuroscience experts and found all of the LLMs to be outperforming the neuroscientists. While LLMs averaged 81% accuracy, humans averaged 63%, which increased only to 66% at the highest level of human expertise. Meanwhile, training an LLM on neuroscience literature improved its accuracy to 86%.

“We suspect it won’t be long before scientists are using AI tools to design the most effective experiment for their question. While our study focused on neuroscience, our approach was universal and should successfully apply across all of science.”

– Senior author Bradley Love, a professor at UCL Psychology & Language Sciences

AI, according to Cambridge researchers, has a clear advantage in terms of predictive modelling and data analysis. When provided with timely data in terms of volume, variety, and veracity, it can optimize costs and supply chains, design high-performing products faster, and respond to market fluctuations in real time.

“Ignoring generative AI in corporate strategy is no longer viable,” said the study co-authors.

This is just the tip of the iceberg, though, as other studies have found AI to even be excelling in fundamental language mechanics⁶ but lacking in thematic consistency when it comes to assay evaluation, outperforming human experts in recognizing pain in sheep⁷ using the same visual information, and matching or best dermatologists⁸ in image‐based skin cognition diagnosis.

Investable Angle: Gemini Adoption through Alphabet Inc. (GOOG +2.51%)

When it comes to investing in the power of AI, Alphabet is a worthy option, which has been leading AI breakthroughs through Google DeepMind and Google Research.

Recently, Google DeepMind and an AI-powered educational technology company, Eedi, released exploratory research⁹ showing that human-in-the-loop AI tutoring outperforms human-only support.

The trial took place in five UK secondary school classrooms, where core instruction was delivered by LearnLM, Google’s generative AI model fine-tuned for pedagogy. They found the human-AI team to be as effective (93%) at helping students immediately fix a mistake they made as a human tutor alone (91.2%). The team was just as good at helping students resolve their underlying misconceptions.

When measuring “knowledge transfer,” which refers to how tutoring on one problem affects a student’s ability to solve a new one, a human tutor alone improved learning by 4.5 percentage points, while the human-AI team boosted it by 10 percentage points.

“These findings mark a milestone for responsible, safe, and effective AI in education. The next step is to scale this from an exploratory pilot to a large-scale trial.”

– Irina Jurenka, Research Scientist at Google DeepMind

Meanwhile, the tech giant’s multimodal AI model Gemini has been making several advances, achieving gold-medal-level performance at the 2025 International Collegiate Programming Contest (ICPC) World Finals, following a gold-medal win at the International Mathematical Olympiad. Google’s flagship AI app now boasts over 650 million monthly active users.

So, Google is deeply involved in AI innovation, and this has helped its stock surge almost 54% this year, now trading just above $293. The $3.5 trillion market cap company also pays a dividend of 0.29%.

Recently, it reported revenue of $102.35 billion for Q3 2025, driven by solid momentum in its cloud business, which benefited from strong AI demand. The company is now planning to increase its capital expenditure to between $91bln-$93bln, after raising expectations from $75bln to $85bln earlier this year. Most of its expenses go toward infrastructure like data centers.

Final Thoughts

AI has taken the world by storm, but it is no longer a standalone innovation. Instead, it serves as a connective tissue linking a growing number of transformative technologies. As mentioned above, it is helping us decode the mysteries of the human microbiome, forecast solar storms, and even outperform experts in science and medicine. Yet these breakthroughs represent only a small part of how deeply AI is influencing nearly every frontier of human endeavor.

As it converges with big data, biotechnology, cloud computing, robotics, and quantum science, AI is accelerating the pace of discovery and enabling integrations that are creating smarter systems and a smarter world.

Click here to learn all about investing in artificial intelligence.

References

1. Mussavi Rizi, M., Fernández, D., Kramer, J. L. K., Saigal, R., DiGiorgio, A. M., Beattie, M. S., Ferguson, A. R., Kyritsis, N., Torres-Espín, A., & TRACK-SCI. Modeling trajectories of routine blood tests as dynamic biomarkers for outcome in spinal cord injury. npj Digital Medicine 8:470 (2025). https://doi.org/10.1038/s41746-025-01782-0
2. Lee, S., Raza, S., Lee, E.-J., Chang, Y., Ryu, S., Kim, H.-L., Kang, S.-H., & Kim, H.-N. Metagenome-assembled genomes reveal microbial signatures and metabolic pathways linked to coronary artery disease. mSystems e00954-25 (2025). https://doi.org/10.1128/msystems.00954-25
3. Dang, T., Lysenko, A., Boroevich, K. A., & Tsunoda, T. “VBayesMM: variational Bayesian neural network to prioritize important relationships of high-dimensional microbiome multiomics data.” Briefings in Bioinformatics 26(4), bbaf300 (2025). https://doi.org/10.1093/bib/bbaf300
4. Sinha, A., Dhuri, D., Vasanth, R., Hanasoge, S., et al. “A Multimodal Encoder–Decoder Neural Network for Forecasting Solar Wind Speed at L1.” The Astrophysical Journal Supplement Series 258(2): 1–? (2025). https://doi.org/10.3847/1538-4365/adf436
5. Luo, X., Rechardt, A., Sun, G., Nejad, K. K., Yáñez, F., Yilmaz, B., Lee, K., Cohen, A. O., Borghesani, V., Pashkov, A., Marinazzo, D., Nicholas, J., Salatiello, A., Sucholutsky, I., Minervini, P., Razavi, S., Rocca, R., Yusifov, E., Okalova, T., Gu, N., Ferianc, M., Khona, M., Patil, K. R., Lee, P. S., Mata, R., Myers, N. E., Bizley, J. K., Musslick, S., Bilgin, I. P., Niso, G., Ales, J. M., Gaebler, M., Ratan Murty, N. A., Loued-Khenissi, L., Behler, A., Hall, C. M., Dafflon, J., Bao, S. D. & Love, B. C. Large language models surpass human experts in predicting neuroscience results. Nature Human Behaviour 9, 305–315 (2025). https://doi.org/10.1038/s41562-024-02046-9
6. Bouziane, K. & Bouziane, A. M. AI versus human effectiveness in essay evaluation. Discover Education 3:201 (2024). https://doi.org/10.1007/s44217-024-00320-6
7. Feighelstein M., Luna S.P., Silva N.O., Trindade P.E., Shimshoni I., van der Linden D., & Zamansky A. “Comparison between AI and human expert performance in acute pain assessment in sheep.” Scientific Reports 15(1):626 (2025). https://doi.org/10.1038/s41598-024-83950-y (PubMed)
8. Ma, X., & Li, Z. Artificial intelligence in dermatology: a review. International Journal of Dermatology and Venereology 7, 227–235 (2025). https://doi.org/10.1097/IJD.0000000000000000
9. Gomes, B., McKee, K. R., Veerubhotla, A. S., Modi, A., Rysbek, A., Huber, A., Wiltshire, S., Gillick, D., et al. “AI tutoring can safely and effectively support students: An exploratory RCT in UK classrooms.” arXiv pre-print (Nov 2025). Available at: https://storage.googleapis.com/deepmind-media/LearnLM/learnLM_nov25.pdf