Blog: Our AI Ethics Pledge at Akademy.ai
- I will aim to apply AI for the benefit of humanity at all times.
- I will respect every human’s privacy as if it were my own.
- I will strive to share the knowledge I develop with others.
- I will seek to be a positive role model for others to emulate.
- I will think carefully about the impact of my models and strive to prevent harm at all costs.
- I will do my best to mitigate the impact of bias in my work.
- I will prioritize human concerns over technological concerns.
- I will work to reduce human inequalities.
- I will keep my word.
If you agree on these principles, you can sign the oath here.
The reasons behind it:
Artificial intelligence has the potential to become a key component of the world’s present and future, shaping the way our society operates and making the way towards the fourth industrial revolution. Such disruptive technology has the power to self-improve without the need for human input. In the long term, it will have the capacity to outsmart us, to the point that some experts argue that it is the last invention that humanity will need to make.
AI is likely to be the best or worst thing to happen to humanity
— Stephen Hawking
What we make of this potential future vastly varies according to how and why we use this technology. While Artificial Intelligence (A.I.) can be used to synthesize new proteins or cure cancer, it also poses a deep thread when placed into the wrong hands. Many experts have called for the ban of AI towards an arms race and facial recognition in cities like San Francisco.
The inevitability AI
One might ask:
If AI has the potential to lead us towards a dystopian scenario in which human life ceases to have meaning, why not ban it all together?
After all, humans could agree to stop using AI systems while continuing to prosper, right? Yes and no.
A set of humans could agree to ban AI tomorrow, much like San Francisco is seeking to narrow its applications in the facial recognition field. The problem with this revolutionary technology is that it has a huge upside for those who use it, specially if such nations or companies are technologically behind.
The U.S. could also agree to ban cell phones tomorrow due to their impact on killing bees but it is highly unlikely that the rest of the world would follow suit. Cell phones give us a competitive advantage, and we’re not willing to give it up -despite the long term implications.
The same is true with AI. Unless there is a global ban or code in which everyone agrees to follow, regulating it for a particular case of individuals leads us to game theory scenarios in which the other parties are incentivized to lie and continue developing such technology on their own with their standards.
Going back to the Origins of Ethics
While it is likely that our prehistoric ancestors had some sort of moral compass, one could make the argument that the roots of ethics as we know it today were set in the Ancient Greece. More in particular, we can trace them back to the ancient Greek philosopher Socrates. He introduced the concept of teaching ethics and acceptable standards of conduct in 400 B.C. and has had a profound and lasting impact on the course of Western philosophy and history ever since.
The understanding of mathematics is necessary for a sound grasp of ethics.
— Socrates, 470–399 BC
He believed virtue was found primarily in human relationships, love and friendship, not through material gains. With his dialectic method his pursuit of the purest forms of knowledge led him to establish a code to have meaningful conversations with people about basic principles, setting aside the parts that set people apart and focusing on the baselines in which they could agree or disagree upon.
The first moral code (of medicine)
A contemporary of Socrates, the Greek physician Hippocrates created the Hippocratic Oath. In its original form, it required a new physician to swear to uphold specific ethical standards. For the first time in history, physicians had to agree to stick to certain principles such as medical confidentiality and non-maleficence.
While we cannot go back and ask Hippocrates why he did it, it is not hard to imagine a scenario in which he realized the power that the contemporary physicians of his time had in their hands. To take or give back a life with the stroke of a scalpel requires a great responsibility and the code allowed for a standard of practice and a set of red lines that could not be crossed.
The Academic club was exclusive and not open to the public, perhaps following the pursuit of the Aristos -greek word for best-. Yet, there was not a membership fee, making it accessible for those that could be deemed worthy. It did not seem to be a formal school, with a clear distinction between teachers and students, or even a formal curriculum.
The Akademy fostered research not just in philosophy narrowly conceived but in a wide range of endeavors that today would likely be called mathematical or scientific. Among its most notable disciples, it included the matemathicians Theaetetus and Eudoxus of Cnidus, the philosopher Aristotle, and was open to women, with two of them being known to have studied at the Akademy: Axiothea of Phlius and Lasthenia of Mantinea.
With this principle in mind, we are launching the akademy.ai. A 3 months bootcamp on Artificial intelligence that combines both a practical approach with a solid ethical foundation. In many ways, it was essential to go back at the original idea of Plato’s akademy and update some of its most innovative ideas into our modern world. We aim for our school to be at the forefront of AI education, rewarding knowledge and meritocracy on top of any other metric. Our world is made of a diverse set of individuals, and we currently do not have an equal set of opportunities in life. Our goal with this program is also to empower the best people from any kind of background, sex, race, sexual identity, age and region to prepare for the future of work.
Also, we have revisited the Hippocratic oath, updating it to the standards that we believe should be set for the modern AI Engineer. The Draper University Startup Hero pledge was a relevant source of inspiration in this process, since I had the opportunity in summer ’17 of experiencing its impact and the seriousness in which each of its students upheld the oath.
Artificial Intelligence has the potential to fundamentally transform the way our society operates, creating one of the biggest paradigm shifts that our society has experienced. And yet, there is a clear gap between the number of professionals out there right now and the ever-growing needs of companies. Currently, AI engineers benefit from the perfect recipe for a substantial paycheck and numerous job opportunities: a hot field and high demand for scarce talent.
But who is positioned to secure a job offer in this field in the future? Very few universities offer AI degrees and even those that exist cannot adapt fast enough to this rapidly-evolving field which is overwhelmed by scientific publications, discoveries and improvements every day, both from academia and corporations such as Google, Facebook or NVIDIA.
There is a global pool of talent ready to learn AI and yet, most of them cannot afford the time and money that it takes to go through a 4-year university program which will become obsolete halfway through.
Our goal is to change this equation by fundamentally decreasing the costs to get technical in this field, all while providing the right pace and environment to get you started as soon as possible.
What we ask for in return? That you put every drop of passion and hard work towards achieving this goal and that you use what you learn to make this world a better place.
The fight against bias
One of the biggest and hardest problems to solve in the field is the bias of AI applications. Besides the black-box calculations of the neural networks, the problem is also related to the models which are populated by human-generated data. While we strive to tackle that problem by updating our daily behaviour, we also seek to use balanced datasets when possible, all while making our academy an inclusive place for everyone where human bias and judgements have no place in our daily strive for knowledge and excellence.
The need for diversity
One of the key ways to solve both for bias and equal opportunities is to ensure that our academy is a home for diversity. From gender to nationality, age and background, the combination of unique individuals is one of our core ingredients for success. It is the richness of perspectives and skills that every member of a team contributes which enables the team as a whole to create high-quality, impactful projects.
AI for Good
Within the increasing potential of AI to disrupt society, our goal is to ensure that the transformation is a positive one. In medicine, transportation, energy and many other fields, AI can yield a significant productivity increase in fields that are essential towards human development. And yet, due to the scalability of AI, it is also a risk if not used for the right causes or if its positive primary implementation has dangerous secondary applications. Therefore, the fellows at our AI Saturdays track commit to creating positive social impact, open source projects in exchange for their accessible education.
In addition, AI can become a big problem if the end does not justify the means. Datasets must be obtained in a fair way while respecting the privacy of individuals and local laws. Image and data labelers should receive fair salary compensation well beyond the “Amazon Mechanical Turk” industry standard in order to reward a job well-done and increase a sense of responsibility for accurate labeling. We also need to start a more active conversation about the negative implications of AI for the future of jobs and in which ways we are responsible for solving the problem that AI will cause for workers. In essence, AI can replace jobs quickly because of the high productivity it leads to and the low costs associated with it, while humans need time and motivation to learn a new skill and money to be compensated for it. However, we are a society of humans so human rights and well-being are our first priority, and we therefore push our students to create end-to-end solutions that leave no human behind.
So, are you ready to sign the pledge?