AI should serve human interests without conflict from "summary" of Human Compatible by Stuart Russell
The fundamental idea is that AI systems should reliably and effectively pursue our objectives without causing harm. This principle is straightforward and uncontroversial, but it has profound implications. It means that AI should be aligned with human values - that is, it should do what we want it to do. If we specify our objectives as "maximize the number of smiles," then the AI should work to maximize the number of smiles. If we specify our objectives as "cure cancer," then the AI should work to cure cancer. If we specify our objectives as "make as much money as possible," then the AI should work to make as much money as possible. The challenge is to design AI systems that are provably aligned with human interests under all circumstances and to ensure that they remain so as they become more intelligent. The AI's objectives must be well-specified. They must capture our true desires and values, as opposed to some distorted version of them. They must take into account the full range of consequences of the AI's actions, not just the immediate effects. And they must be robust, in the sense that they are not easily subverted or undermined by unforeseen events or adversaries. In practice, achieving alignment means solving a difficult technical problem, which is how to design an AI system that can reliably learn our ob...
Similar Posts
The boundaries between humans and machines are blurring
As technology continues to advance at an unprecedented rate, the line that once clearly separated humans from machines is becom...
AI has the potential to enhance equality and social justice
AI holds the promise of helping to level the playing field when it comes to equality and social justice. One way in which AI ca...
Superintelligent AI may surpass human intelligence
The idea that AI might one day surpass human intelligence is a concept that has fascinated and terrified people for decades. In...
The future of AI relies on humancompatible design
For AI systems to be truly beneficial to humanity, they must be designed in a way that aligns with human values and goals. This...
Value learning algorithms development
Value learning algorithms development concerns the task of designing algorithms that can learn what humans value, with the aim ...
Sentient machines can improve decisionmaking processes
Sentient machines have the potential to revolutionize decision-making processes in ways that were previously unimaginable. By h...
The ethical dilemmas of AI mirror those of human decisionmaking
The ethical dilemmas of AI are not so different from those faced by humans when making decisions. When machines are programmed ...
Society must prepare for automation
The rise of automation is rapidly reshaping the landscape of work in our society. As machines become more sophisticated and cap...