44th Reflection: The Human-AI Alliance in the Swarm Above.

A groundbreaking superpower is nearly upon us—an unprecedented ability that no human has ever possessed: the capability to enhance our limited vision through cognitive machines, enabling us to observe the entire world, everywhere, all at once!

Technically speaking, every house, ship, car, cow, and tree on Earth is visible at an average resolution of around 50 centimetres per pixel, with satellites imaging the entire world daily.

However, most of this satellite data goes unseen by human eyes.

To examine all these images, over 33 million full-time employees would be needed—a workforce that doesn't exist anywhere in the world! [1].

At a lower resolution, 70 cm per pixel, if we generated data for all lands covering 29.2% of the Earth's surface or 150 million sq km (58 million sq mi), we'd be bringing ~300 trillion pixels (over 150 TB of data) down to Earth every single day.[2]

If it took an "impossible 5 seconds" for a human to examine each square kilometre of imagery, a single person would need 80 years to view the entire planet.

To complete the task in 24 hours before a new batch of data arrives, it would take 10,000 people—an utterly impractical endeavour!

Fortunately, the right tools can make the impossible possible.

In this quest to derive insights on a global scale, algorithms can make all the difference.

Commandeered by the expertise of the Geospatial Data Analyst [soon to be the AI-first Geospatial Data Analyst], Computer vision algorithms can be leveraged to process terabytes of satellite data to produce a searchable database of the whole world, updated daily, as well as visualizations and predictive models of economic activities, changes in vegetation, migrations of animals and people, the effects of climate change, and so on.[1]

On this note, I welcome you to AI Fluency Friday—my weekly response to the leadership challenge inspired by Eric Schmidt, Daniel Huttenlocher, and the late Henry Kissinger.

Today, in the 44th edition, we'll—(1) explore the unfolding reality of AI-powered Earth observation and—(2) outline the opportunities arising from the new branch of the global digital economy, driven by tens of thousands of satellites being launched into low Earth orbit between 2024 and 2030.

AI Fluency Friday, based on the traditions of immersion , pragmatism , and active imagination ; dates back to March 24, 2023, when the first edition of AI Fluency Friday was published.

As mentioned earlier, this reflection is inspired by the over 107,000 spacecraft applications (satellites) filed since 2020 with the FCC, ITU, and sometimes announced in mainstream media, all scheduled to enter Low Earth Orbit by 2029.[3]

SpaceX, the dominant player in this new space economy, has launched 6,738 satellites, with 6,229 in orbit, and 6,156 of them in operation [4], and by mid-2027, the company plans to launch about 42,000 Starlink satellites into low Earth orbit—12,000 Starlink satellites have been permitted by the U.S. Federal Communications Commission (FCC), and the paperwork for 30,000 additional satellites has been filed with international regulators.[5]

For context, as of November 7, 2022, only 14,450 satellites have been launched throughout history, with 6,800 currently active, according to the European Space Agency (ESA).[5]

The immediate consequence of this spike in the population of satellites is "disruption of the status quo," in global connectivity, and a new paradigm in earth observation, resulting in significant economic opportunities for businesses and individuals.

Concerning economic opportunities, approximately 2.7 billion people globally lacked internet access as of 2022, with significant variations across regions, about 89% of people in Europe and over 80% in the Americas had internet access, while only around 40% of people in Africa and 61% in Asia had reliable internet connections. [10]

Fortunately, as seen above in Figure 3, the cost of a satellite phone (purple bar) and the price per minute of a call via satellite (green bar) were both 15x and 40x higher than terrestrial cellular in 1998 respectively are now 1:1 i.e., without any special modification, a smartphone in a remote location can access internet access directly to a satellite at the same cost as mobile networks. This is made possible via satellites with Direct-to-Cell (D2C) capabilities.

At the time of this writing, there are more than 113 Starlink satellites with D2C capabilities, following the launch of 20 Starlink satellites to orbit including 13 with D2C capabilities, on the 3rd of July, 2024, and 21 Starlink satellites including 13 with D2C capabilities on the 28th of July, 2024.

The journey to SpaceX dominating low earth orbit with 42,000 satellites by 2027 is on, and with the impending possibility of the SpaceX Starship project succeeding, the relatively low cost of launching satellites into orbit via reusable Falcon 9 rockets will crash further, improving the chances for 2.7 billion people to access the internet and participate in the global digital economy via cost-effective satellite technology.

Clearly, the global connectivity revolution is on track, with earth observation next in line to witness a major disruption.

Shifting our focus to earth observation, hiring 33,000,000 employees to examine hundreds of terabytes of daily satellite images of the entire world is not feasible.

However, the opportunity exists for a new breed of data analysts, specifically AI-native geospatial data analysts who can work with vision models to extract insights from satellite images.

In August 2024, working with raw geospatial data requires specialized knowledge and the application of advanced mathematics to conduct necessary tasks, such as geospatial alignment of data layers.[6]

Nevertheless, there is an expectation of the “ChatGPT 3.5 moment” of Earth Observation [2], a point at which we will have attained true democratization of Earth Observation.

The democratization of Earth observation refers to making high-quality satellite data accessible to a wider range of users, including businesses, governments, and individuals.

This involves providing affordable, user-friendly tools and scalable cloud-based platforms, enabling non-experts to analyze imagery and derive insights without deep technical expertise.

Advances in AI and visual language models (VLMs), and their adoption in earth observation have significant implications for increased investment in (1) Sustainability, (2) economic losses from natural disasters, and (3) the financial exposure of companies to climate risks. Let's briefly discuss these implications.

1—87% of business leaders are increasing their sustainability investments in the next two years.[7]

Impact: Enhanced Decision-Making and Accountability

a—Enhanced Data Access: With democratized Earth observation, businesses of all sizes can access high-quality, real-time environmental data, which can inform more effective sustainability strategies and investments.

b—Precision Monitoring: Companies can monitor their environmental impact with greater precision, enabling them to adjust operations to meet sustainability goals, reduce waste, and improve resource efficiency.

c—Transparency and Reporting: Visual language models can improve the transparency of sustainability practices, providing detailed reports and visual evidence that can be used for internal assessments and external reporting to stakeholders and regulatory bodies.

2—Damage from global natural disasters in 2023 totalled USD 380 billion in economic losses?– with only USD 118 billion of these losses covered by insurance.[8]

Impact: Improved Preparedness and Response

a—Early Warning Systems: Advanced Earth observation capabilities can improve early warning systems for natural disasters, giving communities and businesses more time to prepare and mitigate damage.

b—Disaster Response: Enhanced imagery and data can aid in the rapid assessment of disaster-stricken areas, optimizing the allocation of resources and efforts in disaster response and recovery.

c—Insurance Models: Insurers can use the data to refine risk models and offer better coverage options, potentially increasing the proportion of insured losses and reducing the economic impact of disasters on businesses.

3—By the 2050s, 90% of the world’s largest companies will have at least one asset financially exposed to climate risks such as water stress, wildfires, or floods.[9]

Impact: Risk Management and Resilience Building

a—Risk Identification: Visual language models can help identify and map areas at high risk of climate-related events, allowing companies to take proactive measures to protect assets.

b—Resilience Planning: Businesses can use this data to design and implement resilience strategies, such as infrastructure improvements and adaptive measures tailored to specific climate risks.

c—Investment Prioritization: Companies can prioritize investments in regions or assets that are most vulnerable, ensuring that resources are allocated effectively to mitigate future risks.

Overall, the democratization of Earth observation using visual language models empowers businesses, governments, and communities with actionable insights, leading to more informed decisions, enhanced preparedness, and a stronger collective ability to address sustainability and climate challenges.

At this point, it is necessary to recall that this impending commoditization of intelligence and the 'ease of performing analytics' are all at level 1 of OpenAI's AGI maturity index.

Consequently, human creativity, critical thinking, imagination, and the curation outputs of geospatial solutions will remain crucial skills.

At any rate, for a market anticipated to increase from the 2024 size of $89.81 billion to as much as $141.9 billion by the end of 2028, and $262.73 billion by 2032 [11], with a CAGR of 14.4% during the forecast period (2024-2032), there is a wide skills gap limiting the market.

In a Geospatial World Annual Readership Survey, around 84% of 2,000 surveyed industry professionals acknowledged that there was a skill gap in terms of industry requirements and availability of workforce. [12]

Though our reference to Geospatial data analytics has mostly focused on satellite imagery, note that geospatial analytics includes the gathering, manipulation, and visualization of different types of geospatial data generated from GPS, location sensors, drones, and smartphones.

The full spectrum of possibilities in geospatial data analytics in the age of AI is appreciated by no other than the military-industrial complex which can deploy more satellites, and derive far more insights than was ever possible.

At any rate, our focus is on the civilian space, and only four years ago, the World Economic Forum's Future of Jobs Report illuminated the dual nature of AI's impact, positing that by 2025, 85 million jobs may be displaced by automation, an impressive 97 million new roles are projected to emerge, reflecting a shift in the division of labour between humans, machines and algorithms.

However, given the progress of the last 20 months, and the impending development of the "new AI workforce," along the spectrum of the AI maturity index, the matter of future jobs is in a state of flux, with a school of thought arguing for the emergence of a 'useless labour class,' and other groups hedging their bets on the possibility of Jevons Paradox leading to more jobs than there are humans to do them, an age of true abundance.

No matter one's opinion about the future of knowledge work, insights from the 2023 Future of Jobs Report are worth considering.

Specifically, the report indicates that the largest job creation and destruction effects will come from environmental, technological and economic trends.

Climate change adaptation and the demographic dividend in developing and emerging economies also rated high as net job creators.

Recalling our earlier reflection on the importance of democratized earth observation for (1) Sustainability, (2) economic losses from natural disasters, and (3) the financial exposure of companies to climate risks, as well as the skill shortage in earth observation, we can confidently expect that some of the 'safe jobs' of the future will be that of the AI-native geospatial analyst.

In closing today's reflection, recall the last time global GDP increased by 10x, it took 190 years, from 1820 to 2010. And with the promise of AI, another 10x is possible within the 21st century.

However, some major concerns remain, such as where will Africa, the continent that will have 1/3rd of the world's young people by 2050, fit into the new world of AI/AGI?

Finally, having drawn attention to the global connectivity revolution, and more importantly, the impending democratization of earth observation, this article has highlighted an opportunity, no matter how little, which can be leveraged to address the BIG question of how Africa might maximize its remarkable demographic dividend, YOUTH.

In the coming months, I will be contributing to efforts to achieve the following research objectives for developing economies (i.e., Africa) in the new space economy:

1. Assess the current state of satellite technology education and skills among African youth across different regions of the continent.
2. Identify key areas within the satellite-based economy where African youth can make significant contributions, considering regional variations and needs.
3. Develop preliminary strategies for integrating satellite technology and geospatial data analysis into relevant curricula across African educational institutions.
4. Explore potential applications of satellite technology for addressing Africa-specific challenges in areas such as agriculture, urban planning, environmental conservation, and disaster management.
5. Conduct a preliminary investigation into the economic impact of increased African participation in the global space economy, including potential job creation and entrepreneurship opportunities.
6. Examine the role of satellite technology in bridging the digital divide and its potential impact on youth empowerment and economic inclusion across Africa.

This has been the 44th edition of AI Fluency Friday (#AIFF). I am Orakwe John, MBA , and I collaborate with global businesses and institutions to envision and build the future.

If you're ready to be part of the growing list of AI-powered businesses, business leaders, and professionals with whom I'm crafting AI success in 2024, reach out, drop a message, and let's forge a legacy together. See you again soon in the 45th Reflection!

References:

[1] Russell, S. (2019). Human compatible: AI and the problem of control. Penguin Uk.

[2] https://satellogic.com/2024/04/24/the-future-of-earth-observation-lies-inside-visual-language-ai-models/

[3] https://www.agi.com/resources/videos/107-000-planned-satellites

[4] https://planet4589.org/space/con/star/stats.html

[5] https://www.space.com/spacex-starlink-satellites.html

[6] https://www.ibm.com/topics/geospatial-data

[7] https://www.techtarget.com/sustainability/feature/How-does-climate-change-affect-businesses-Financial-impacts

[8] https://www.zurich.com/knowledge/topics/climate-change/how-climate-change-will-impact-business-everywhere

[9] https://www.bloomberg.com/news/articles/2023-07-12/record-heat-is-burning-investors-and-they-don-t-know-it

[10] https://robertsmith.com/blog/global-digital-divide/

[11] https://www.marketsandmarkets.com/Market-Reports/geospatial-analytics-market-198354497.html

[12] https://www.fortunebusinessinsights.com/geospatial-analytics-market-102219

[13] https://www.forbes.com/sites/forbestechcouncil/2024/03/12/the-future-of-work-embracing-ais-job-creation-potential/

[14] https://interactive.satellitetoday.com/via/november-2023/how-ai-and-ml-are-supercharging-earth-observation/

[15] https://saiwa.ai/blog/computer-vision-in-satellite-imagery/

[16] https://www.ark-invest.com/white-papers/platforms-of-innovation