TL;DR:
- Maniac, a post-WWII computer, performed 10,000 calculations/second for modeling thermonuclear explosions and weather.
- Today’s supercomputers excel in weather prediction but still devote capacity to weaponry.
- Numerical weather prediction’s success amounts to $162bn/year, benefitting various sectors globally.
- Machine learning and AI enhance weather forecasting, revealing new insights and attracting innovative entrants.
- To optimize possibilities, healthy competition must coexist with data-sharing and infrastructure preservation.
- Combining AI with number-crunching tackles climate change and improves early warning systems.
- Accessible weather forecasts can save lives and livelihoods with modest investment and political will.
Main AI News:
Maniac, a groundbreaking computer developed at Princeton post World War II, boasted an astounding capacity of 10,000 calculations per second. This extraordinary computational power was primarily utilized for two pivotal purposes: simulating thermonuclear explosions and predicting Earth’s weather patterns. The significance of these applications was unparalleled in the eyes of the machine’s creators.
Fast forward to today, and the disparity is staggering. The swiftest computers of our time can accomplish in a mere hour what it would have taken Maniac the entire 13.8 billion years of the universe to achieve. Despite this leap in capabilities and ambition, modern supercomputers still allocate a considerable portion of their resources to matters of weaponry and weather. While their contributions to H-bomb design instill a lingering sense of unease in most people’s lives, their weather forecasting work positively impacts almost every facet of our existence.
Research conducted by the World Bank and other reputable institutions values the benefits of numerical weather prediction (NWP) at a staggering $162 billion annually. This success is evident to any contemporary farmer or military strategist, as well as in the fabric of our daily lives. The weather-related icons on our smartphones signify a reliance on forecasts to plan our activities, and trusting a forecaster’s advice has become a pragmatic choice rather than a leap of faith.
The integration of machine learning and various forms of artificial intelligence (AI) stands to propel advancements even further. Today, NWP supercomputers rely on present conditions, physical laws, and rules of thumb to predict tomorrow’s weather, necessitating trillions of calculations at a high resolution. Yet, machine-learning systems trained on historical weather data can now rival these forecasts in certain respects. If the progress in AI across other fields serves as any indication, this is just the tip of the iceberg.
Moreover, AI appears capable of unearthing aspects of weather behavior that traditional NWP methods struggle to reach through calculations alone. With lower costs compared to traditional methods, AI solutions are attracting new players to the weather industry. These newcomers are expected to bring highly tailored products to cater to customer needs, while also presenting fresh ideas that can unlock new markets.
To seize these possibilities, three crucial actions must be taken. First, it is essential to safeguard the fundamental infrastructure by fostering healthy competition. Governmental entities predominantly dominate NWP, putting substantial effort into assimilating worldwide weather observations to ensure consistent representations in their models. These high-value forecasts can be sold to niche markets to offset the costs.
However, to achieve the best results, AI systems need to be trained using data from these representations, potentially undercutting some existing forecasters’ offerings. A delicate balance must be found, ensuring that generous data sharing with new entrants does not jeopardize the established systems that provide crucial data sets for AI and the world’s reliance, at least for the time being.
The second action involves harnessing the synergy between AI and number-crunching to address the pressing issue of climate change. Current climate models lack the resolution of weather forecasts, but advancements in hardware designed for AI systems could bridge this gap. AI also holds the potential to identify patterns in the models’ projections, making them more informative and accessible to non-experts.
In the present, however, improved access to weather information is urgently needed. A mere 24-hour warning of a destructive weather event could reduce damage by 30%, as highlighted by the Global Commission on Adaptation in 2019. A targeted investment of $800 million in early-warning systems for developing countries could prevent annual losses ranging from $3 billion to $16 billion. The World Meteorological Organization has rightly prioritized “Early Warnings for All” by 2027, considering that three out of four people worldwide own mobile phones, yet only half of their countries possess disaster warning systems.
Addressing this issue doesn’t require groundbreaking breakthroughs; rather, it demands modest investment, meticulous planning, focused discussions, and unwavering political determination to overcome institutional barriers. While it may not ignite the world like Maniac’s legacy, this effort can undoubtedly save thousands of lives and safeguard millions of livelihoods.
Conclusion:
The marriage of AI and numerical weather prediction has revolutionized the weather forecasting market. The enormous potential of AI-driven weather predictions promises improved accuracy, new market opportunities, and better climate change understanding. However, stakeholders must navigate the delicate balance of sharing data with new entrants while maintaining the integrity of existing forecasting systems. Embracing this technological shift and investing in early-warning systems will lead to significant improvements in disaster preparedness, saving lives, and protecting livelihoods worldwide. Business players need to capitalize on AI advancements to stay competitive in this rapidly evolving market.
