In a world awash with buzzwords like blockchain, metaverse, and AI, it’s easy to get caught up in the hype. Yet behind the noise, some companies are applying these technologies in practical and transformative ways. Wolfram Research, a leader in computational intelligence and scientific innovation, has been quietly driving advancements in fields like data science, machine learning, and AI for decades. We spoke with Jon McLoone, Director of Technical Communication and Strategy at Wolfram Research, to explore how his organization approaches the intersection of AI and real-world problem-solving.
Wolfram’s Unified Approach to Computation
For more than 30 years, Wolfram Research has built a reputation as a leader in technical computing. As McLoone describes it, Wolfram’s value proposition lies in its ability to tailor its powerful computational technologies to solve specific problems across industries. “We’re not solving the same problem for every customer,” McLoone explains. “What our customers have in common is that they’re doing something innovative.”
At its core, Wolfram focuses on symbolic computation, meaning it doesn’t just rely on large data sets but also on deep mathematical models and structured representations of real-world entities. From social network analysis to biosciences, and actuarial science to financial computations, Wolfram’s approach is rooted in mathematical rigor, providing businesses with tools to solve complex problems.
The Role of AI in Wolfram’s Computational Framework
While generative AI dominates headlines, Wolfram sees AI as just another tool in the broader realm of computation. McLoone emphasizes that AI, particularly generative models, is good at making connections and producing plausible results, but it lacks the precision of symbolic AI. “Generative AI is fluent but unreliable,” McLoone points out, explaining that while it can generate likely answers, it often lacks the ability to deliver truly accurate solutions, especially in fields that require precise calculations.
He offers an analogy: imagine modeling the trajectory of a thrown ball. A generative AI might observe thousands of throws and produce a plausible description of how the ball will move. However, a symbolic AI would model the situation using differential equations, taking into account factors like mass, air resistance, and gravity. “If you ask it how the ball will behave on Mars, it’ll give you an accurate answer,” McLoone adds.
Combining AI with Human Intelligence and Symbolic Reasoning
Wolfram’s approach to AI involves combining the strengths of human intelligence, symbolic reasoning, and AI to address business, scientific, and engineering challenges. AI can act as a bridge between human language and structured data, making it easier to interpret and act on complex information. For example, Wolfram recently worked on a project involving medical records that existed in various formats, from handwritten notes to digital files. Generative AI was used to classify and structure the data, turning qualitative information into a structured format that could then be analyzed statistically.
By using AI to map unstructured data into something computable, Wolfram can tackle challenges that would otherwise be impossible to address at scale. Whether it’s analyzing hospital records for trends or optimizing the production process at a hypothetical peanut butter cup factory, the company’s ability to model complex systems with computational precision is unmatched.
Practical AI for Business and Beyond
McLoone also discussed how large language models (LLMs) can connect human questions to complex computational models. For example, when asked how changing an ingredient might affect the shelf life of a product, an LLM might give a plausible but generalized answer. However, Wolfram’s models can connect this question to a detailed computational model of the manufacturing process and determine the exact impact of changes on product longevity or production efficiency.
This combination of generative AI and precise computation creates a powerful synergy, allowing companies to make informed decisions that are based on data, models, and mathematical reasoning.
Looking Ahead: The Power of Computation and AI
As AI continues to evolve, Wolfram Research’s approach offers a clear vision of how these technologies can be harnessed for real-world applications. Rather than relying solely on generative models, Wolfram combines symbolic AI, human intelligence, and computational power to solve complex problems across industries. Whether it’s financial modeling, biosciences, or optimizing industrial processes, the ability to ask the right questions and use the right tools remains critical.
Wolfram Research’s philosophy underscores that AI, while a powerful tool, is only one part of the equation. By integrating it with established scientific and mathematical principles, businesses can achieve far more than what AI alone could offer. This balanced approach ensures that innovation continues without falling victim to the hype.
For those interested in seeing Wolfram Research’s capabilities in action, the company will be showcasing its work at the upcoming TechEx event in Amsterdam. It’s a great opportunity to learn how AI and computational models can be tailored to address specific challenges in your industry.
Sources: https://www.artificialintelligence-news.com/news/how-cold-hard-data-science-harnesses-ai-with-wolfram-research/, https://calmatters.org/commentary/2021/08/californias-proposed-new-math-curriculum-defies-logic/
