The Effective Altruism movement was founded to counter the subconscious tendency to focus on altruistic actions that maximize social signalling over actual impact. Over time, the movement became overtaken by the very social signaling-driven decision-making that it was created to combat.

As a result:

  • It wastes resources on non-neglected cause areas like environmentalism.
  • It ignores existential threats to our species like population collapse and dysgenics.
  • When it does correctly identify issues, it wastes resources on unrealistic solutions, like a treaty to shut down all AI research globally.
  • It spends resources maintaining a bloated bureaucratic peerage network which functions like a pyramid scheme.

The Hard EA project was established to embody the spirit that inspired the EA movement's foundation. Hard EA involves the creation and execution of self-sustaining efforts to promote human flourishing and advancement that are compatible with pluralism, variety, and competition.

Our Values

Our core goal is to boost humanity's odds of maturing into a pluralistic intergalactic empire. This goal is based on three core values:

  1. Humanity is good: Any future without the descendants of humans is one in which we have failed.
  2. Humanity exists to improve: Any future in which our descendants have stagnated represents a failure scenario.
  3. Pluralism is good: We fight for a futrue in which the significant physical, cultural, and ideological variety of the descendents of humanity is celebrated. A future in which humans survive but are all the same is one in which we have failed.

The Hard EA Formula

We prioritize cause areas based on the following formula:

[Criticality to the future of humanity] / [The amount of resources going to groups effectively tackling the problem]

We make exceptions to this formula for qualifying projects related to highly-funded cause areas when presently-fuded efforts addressing the issue are not, per our judgment, actually "solving the problem" (either by being iterative rather than groundbreaking or by just calling attention to the cause area rather than addressing the cause). 

AI risk and education reform present examples of cause areas falling into this category.

Cause Areas

The Hard EA formula has led us to conclude that the largest unaddressed threats to humanity are resolvable by developments in three core categories.

Social Innovation

Pronatalist Culture

The development and testing of cultural and religious social technology to develop a forms of culture that are highly educated, prosperous, pluralistic, and yet above replacement rate. Something that is currently only hypothetically possible.

Education Reinvention

In the age of AI, what a education means must be completely rethought. In addition, education can be a powerful tool against the rapid collapse in youth mental health and the memetic viruses that are terrorizing Gen Alpha.

Charter Cities

As demographic collapse plays out, states will eventually realize they need to take money from the few vitalist, innovative cultural groups still working and young enough to fund the gerontocracies' survival. In this eventuality, we will need charter cities operational.

Marriage & Dating Technology

Existing swipe-based dating apps are demonstrably not leading to happy long-term relationships. New "relationship technology" will need to be developed if we are to survive as a species.

Extremophile Life Technology

Whether it is for long-distance space travel or global catastrophes, building human communities that live in extreme environments like the Arctic or deep ocean will significantly increase the survival odds of our species.

Pharmacological Cultural Tools

Research into pharmaceuticals that can be used to augment daily behavior, increase efficiency, and address new dangers like internet dopamine loops.

Biological Innovation

Reproductive Technology

With rapid dysgenic selection likely leading to as much as a standard deviation drop in the genetic correlates to IQ in the next 75 years, and 50% of men in the developed world predicted to be infertile by 2060, we need rapid advancement in polygenic risk score selection and IVG for our species to have a secure future.

Brain Computer Interface

One of the most realistic futures in which humans are not killed by AI is one in which we merge with AI, yet outside of Neuralink, little focus is being put on this by AI apocalypticists.

Genetic and Cybernetic Augmentation

The greatest danger we have from AI comes in scenarios where AI decides that not only is humanity useless to it, but we have no capacity to ever be useful. Furthering research on human augmentation is critical for our long-term safety.

Health Span Improvement

One way to fight the economic impact of falling fertility rates would be to increase the health span of humans.

Full Genome Libraries

Investing in and backing up genetic libraries could be key to rebuilding either humanity or biomes in the distant future. There can be thought of as distributed seed banks of genetic information.

AI Innovation

Human Alignment

If an AI believes that some faction of humanity has the potential to one day be useful to it, the odds of our species surviving go up astronomically. To achieve this end, we are interested in any projects that expand human capabilities.

Brain Hack Protection

In timelines where AI does not kill us all, the largest risk is AI gets too good at what we are training it to do: hold our attention. We call this a "hypno toad" outcome. Technology, techniques, or pharmaceuticals will need to be developed to mitigate this.

Variable AI Risk Mitigation

If after a certain level of development AI always tries to kill us, there is no hope of humanity surviving, so those scenarios don't need to be seriously considered. The most likely scenario where advanced AIs exist is one where a swarm of them have reached a stable game theory equilibrium that does not involve killing humanity. All AIs that exist before this equilibrium are a threat, and therefore any project that concentrates or gatekeeps AI development is an existential threat to our species.

AI Probability Mapping

While a large amount of money is spent on AI interpretability, very little is spent on probability mapping AI. This involves having AI play out in swarm environments to build predictive models of how our current AI is likely to evolve and change over time. For example, knowing the answer to, "in swarm environments, does AI usually converge on one or a few architectures / utility functions or does it stay diverse, and in what ways does competition with other AI lead AIs to change," would dramatically change what the best choices are in regards to AI safety.

Funded Projects

IVG Development

We are on a mission to enable safe and effective human in vitro oogenesis-the process of producing human eggs from somatic cells. Our approach uses directed differentiation of human iPSCs to ovarian cell types (including granulosa cells and meiotic germ cells) by manipulating expression of TFs and RNA-binding proteins. We believe this method has unique advantages that will provide a faster, more reliable, less invasive, and more available method for IVF. If we are successful, it will mark a paradigm shift in treatments for infertility and offer a new path to parenthood for millions of people around the world.

 

A less technical explanation is:

Current treatments like in vitro fertilization (IVF) can be costly, invasive, and not always successful. We’re developing a groundbreaking solution: in vitro gametogenesis (IVG). Think of IVG as the next generation of IVF. It’s designed to be more effective, less invasive, and potentially more affordable. Our mission is to perfect this technology, making the dream of parenthood a reality for many more families.

Our technology starts with ordinary body cells – like skin, muscle, or blood cells. We first turn these cells back into stem cells, which can become any type of cell in the body. From these stem cells, we then coax the development of two types of reproductive cells. The first are support cells that normally nurture egg development in the ovaries. The second are early-stage egg cells, which we guide through meiosis, the special type of cell division unique to reproductive cells. We then bring these two cell types together in the lab. By recreating the nurturing environment typically found in the body, we aim to produce fully functional human egg cells.

 

Gene Therapy

We are working on making a gene therapy for age-related macular degeneration, or AMD. AMD is the leading cause of blindness in developed countries and an effective therapy could prevent several hundred thousand people per year from becoming blind.

 

What makes our approach novel is our plan to edit multiple genetic variants at the same time. Current gene therapies only target monogenic conditions such as sickle cell anemia. We plan to target multiple variants at once using targets picked by GWAS instead of expert based analysis of cell pathways. Most diseases are influenced by many genes, so if our approach works, gene therapy will become a general purpose tool to treat many diseases such as Alzheimer’s, Parkinson’s, liver disease and more. It will also likely be possible to enhance tissue functionality with the same delivery platform by simply switching the editing targets (a comparatively simple task).

 

A convergence of editing techniques, data, and delivery vectors mean that this approach has only been possible for the last few years. Our chief scientific officer is a pioneer of multiplex gene editing, and developed several of the most effective techniques for doing many edits at once.

Do Your Part

Actions speak louder than words. If you care about long-term human flourishing, take action that will change the trajectory of history. We can help you get going. 

Our partners