Assuming you're subscribed to the Abundance agenda, the next question to ask is: Where should we focus our efforts? This brief essay outlines a roadmap of stages on the path to abundance. Then, in later essays, I will outline concrete steps to get us to each stage.

I hope that this roadmap can help students, trainees, and others who want to have an sizeable impact on the world, but are not sure where to focus their energy. I also hope that investors and philanthropists may be inspired by some of the ideas in this roadmap and help fund those who are diligent working towards abundance.

Please note that while the stages here are listed sequentially, we should begin to work on later stages before the earlier stages have been completed.

Painting the vision: What is "abundance"?

Abundance is about making essential goods and services plentiful. Those goods and services include:

• Energy: Higher standards of living almost invariably involves increasing the amount of energy used per capita (e.g. HVAC).
• Housing: Everyone should have their own home, not just as shelter but as a space they control.
• Food: Everyone should have enough to eat, for nutrition and for enjoyment.
• Water: Everyone should have access to clean water.
• Health: Everyone should have the opportunity to a long life.
• Tools and education: Giving people access to tools and knowledge that expand their agency.
• Environment: All of the above need to be accomplished without straining the environment.

Abundance focuses only on the "essentials" because people are different. By making the essentials bountiful, which includes tools and education, everyone will have the agency to pursue their own goals and desires.

A roadmap

Now that we have a concrete vision of the end goal of abundance, we can plan the intermediate steps we need to accomplish to get there. I've ordered the following steps considering both feasibility (technological, economic, and political) and dependency (earlier stages enable the later stages).

Step 0: Enhancing education and tools

The essential prerequisite to everything in this roadmap is to develop better learning tools and systems. Almost every venture is bottlenecked by talent that is able to both think through technical challenges (theory), and execute potential solutions (practice). Although AI and robotics is progressing rapidly, most skilled tasks will be executed by humans for the foreseeable future. And ultimately, skilled human operators will always be needed to manage and evaluate the output of their teams, AI or not (the accountability principle).

Importantly, the problems we are tackling are becoming more complex, requiring a broader range of skills to solve them. Most problems now require a combination of engineering (how to design of new systems that solve problems using knowledge spanning mechanical, electrical, chemical, and software disciplines), manufacturing (how to fabricate), construction (how to assemble), and project management (how to plan, coordinate, and operate). We are also tackling more problems in parallel, which means we need even more talent to be available.

Luckily, we have the technology to make education much more effective and scalable. AI systems can rapidly generate Properly deployed AI systems can scale Bloom's mastery learning techniques to achieve a two-sigma improvement in learning for everyone. AI tutors can properly address student misconceptions and provide the bridges that connect to existing knowledge (the theory of meaningful learning).

We also have the ability to rapid scale praxis. Consumer-grade fabrication tools (3D printers, desktop CNC, laser cutters, vinyl cutters, soldering stations, microcontroller platforms, etc.) and software (Onshape, Fusion 360, etc.) are now readily available. YouTube tutorials have made learning how to use these tools a question of motivation. Now, AI tools lower the barrier even further with their ability to do things like translate jargon and formulating bills of materials for starter projects, making classes like How to Make (Almost) Anything and Machines that Make more self-teachable.

This means that for education, the bottleneck is not in the technology, but in building the systems that bring these technologies together cohesively, so that students are able to rapidly expand their knowledge and capabilities.

Step 1: Deploying general purpose robots

The first major milestone the path to abundance is the development and deployment of general purpose robots. Why? Two reasons:

First, research in robotics is progressing rapidly, which means that coordinated efforts will yield productive results soon. Sergey Levine, renowned robotics professor and co-founder of Physical Intelligence, estimates that in around 5 years, general purpose robots will begin to be useful for real world tasks. Similarly, Jim Fan, NVIDIA's Director of AI and Distinguished Scientist, gave a convincing argument that we are in the "end game" for solving robotics.

Second, robots are a general-purpose technology. They are the equivalent of the computer for the physical world. And robots could not come at a more critical time:

1. The demand for physical labor is growing. Infrastructure projects (energy, factories, etc.) always require a huge amount of workers, and as we accelerate the number of projects being built, labor demand will continue to grow at an increasing pace.
2. The traditional supply of labor (people) is not growing. For a multitude of good reasons, people generally do not desire hard, physical labor jobs. For example, it will be clear that mining and other critical minerals extraction will be required for abundance, but these jobs are notoriously dangerous and inhumane. The ability complete these jobs with robots, managed by people, will be a win-win for both productivity and safety. Many skilled jobs are also risky and should be augmented by robots, such as BL2+ infectious disease research or even surgery, which rely heavily on adherence to strict protocols for human behavior.

Once we have a supply of general-purpose robots, we can really start to scale our efforts in major infrastructure projects, namely energy.

Step 2: Rapidly scaling clean energy production, distribution, and storage

The next step is to scale clean energy production, distribution, and storage. Again, the staging is strategic. We have the technology for producing abundant clean energy: nuclear fission and renewables. And once we have an abundant energy supply, we unlock the economics of a wide array of downstream tasks that enable abundance (including desalination, vertical farming, and direct air capture). The fact that abundant clean energy resolves many bottlenecks at once is the reason energy is one of Demis Hassabis's root node problems).

Although nuclear fission and renewables only produce clean electricity, that is enough to decarbonize non-electrical uses of energy. The most straightforward approach is electrification, where electricity directly substitutes for other sources of energy. But clean electricity can also be used to decarbonize uses where electrification is not possible. For example, hydrocarbon fuels carry a energy density that cannot be matched by batteries. In these cases, abundant clean electricity can make biofuel production economical, which are then used as drop-in carbon-neutral replacements.

It should be stated that production is only one part of the energy milestone. The grid, which makes electricity widely available, needs upgrading. Additionally, grid-scale energy storage systems need to be engineered and deployed in order properly take advantage of increased electricity generation. All of these challenges should be tackled as part of scaling clean energy production.

Energy is essential to every downstream ambition. Therefore, unlocking abundant clean energy is a critical milestone to reach early on.

Step 3: Protecting and remediating the environment

Humanity has made exponential progress to get to where we are, but up to this point, development has also come at the expense of the environment. This may have been a necessary price to pay, but we are reaching a level of technological sophistication where we can continue to grow while protecting the environment.

We must prioritize this step before we use more resources from the environment. While the Earth has enough resources to support all life, including abundance for 10 billion people, how we extract and use those resources matters.

Clean energy will be a major step towards protecting the environment. But other technologies will also be critical to protecting the environment. One of the big challenges, however, is that what those technologies are is not necessarily obvious. Some ideas that may be implemented (of varying degrees of certainty): proper accounting of natural resources; graceful degradation and recycling of materials; remediation robots; ecosystem restoration; de-extinction.

We must prepare for this step early by investing resources into research and exploration because we don't know the answers yet. We need to find strategies that protect the environment and avoid the hubris that has historically punished us (as recorded in Seeing Like a State).

Step 4: Mass production of industrial inputs

Assuming that we are able to properly steward Earth's resources, the next step will be to scale the production of industrial inputs in order to build the factories, machinery, robots, buildings, housing, et cetera that define abundance. A non-comprehensive list of inputs that we need to scale include: semiconductors, batteries, steel, cement, metals, chemicals, etc.

This stage may not have major technical challenges to overcome, but jumping to this step too early will be challenging because energy and labor will be key resource bottlenecks.

Step 5: Accelerating science

Perhaps surprisingly, the previous steps clearly show that our existing technology stack is capable of providing abundance almost every category. This is heartening, since that means coordinated effort can lead us to abundance. Therefore, those steps have been prioritized, to ensure we make timely progress.

Importantly, however, there remains one category of abundance that cannot yet be reduced to an engineering challenge: human health. To date, biology is still too complex and unpredictable for us. In order to develop cures faster, we need to address both fundamental knowledge and technology gaps to better understand the human body.

And beyond human health, accelerating science is still critical, however. Undoubtedly, the technology stack we currently use (e.g. silicon wafers, lithium batteries, etc.) is not the most efficient nor effective. Accelerating scientific discovery will allow us to continue to push the Pareto frontier of what is possible with the resources we have.

Conclusion

I've outlined what I believe to be a plausible roadmap towards abundance. I hope it serves as a useful guide to highlight some of the most important problems (impact $\times$ tractability) to work on for the foreseeable future. I also hope that it resonates with investors and philanthropists, and provides some useful ideas on where to deploy new waves of capital.

In a series of future essays, I plan to go in depth into each of the milestones, and identify concrete steps to get us to each stage. But in the meantime, if others have their own roadmap, or have a detailed playbook for any of the milestones, I'd love to read and discuss them.