“Nature is busy creating absolutely unique individuals, whereas culture has invented a single mold to which all must conform. It is grotesque. ”
―
I have often wondered about the consciousness that a newborn human being has. You had it and I had it, but we cannot recall it. They say that we only have memories, or access to memories, that formed after we learnt a language, our respective mother tongues.
―
I have often wondered about the consciousness that a newborn human being has. You had it and I had it, but we cannot recall it. They say that we only have memories, or access to memories, that formed after we learnt a language, our respective mother tongues.
Recently I read a book titled “The Big Nine.” It is about Artificial Intelligence (AI) and certain concerns about the way it is developing. I have done my fair share of programming, but only in the days of COBOL and xBASE, and so it should not be difficult to see how poor my understanding of Artificial Intelligence is. What I could gather after reading the book is that AI essentially means finding patterns by sifting through humongous amounts of data, using initial instructions on how to do it, what to look for and then applying the knowledge gained to new data sets on new situations. I would like to think that this is strikingly similar to how our our own intelligence, an important facet of consciousness, develops.
All the data and instructions that we accumulate to get our NI (Natural Intelligence) running, comes from our parents, family, teachers and the community. This gives a final shape the idea of 'me' and then, of course, we get weaned from these sources but the foundation of my self remains.
The book I mentioned earlier, expresses serious concerns about the fact that those who are in the forefront of AI development are predominantly white males. This means that their biases in interpreting human data are very likely to get embedded into AI itself. This is the equivalent of a child getting exposed only to the ways of its community to the exclusion of others. Apart from the ways, the data he gets is also limited to that originated by the members of his cultural group. Later on when he encounters people from a different culture, he may feel overwhelmed and even threatened.
The book argues that if this pioneer team were more heterogeneous, the biases may get cancelled out. It is the equivalent of bringing up a child in a multicultural environment.
An AI, after it gets started, may, later in its 'life', encounter data that significantly deviates from the ones that it has trained on and also finds its established methods failing in handling this data. Depending on the kind of intelligence bred into it, it may exhibit behaviours that may be difficult to predict in advance and less than desirable. This wouldn't be the case if it was built up using an inclusive instructions and data.
Extrapolating the argument, it is important that humans too be exposed to different cultures as well as the common culture of a super-group like a nation during the early formative years, to avoid them displaying confused and roguish behaviour later in life.
Such bringing up will make a person more rounded without rough edges and spear points. Such a person will not have the desire to hurt or eliminate an alien behaviour on emotional grounds alone.
But yet, the fact remains that even this well rounded person and his consciousness is merely what he has learnt from his (more inclusive) society and gives no clue whatsoever about what it was like when it emerged into this world. Just as an AI cannot be described just on the basis of its hardware and OS. In fact it emerges only after initial instructions and exposure to the humongous data sets.
All the data and instructions that we accumulate to get our NI (Natural Intelligence) running, comes from our parents, family, teachers and the community. This gives a final shape the idea of 'me' and then, of course, we get weaned from these sources but the foundation of my self remains.
The book I mentioned earlier, expresses serious concerns about the fact that those who are in the forefront of AI development are predominantly white males. This means that their biases in interpreting human data are very likely to get embedded into AI itself. This is the equivalent of a child getting exposed only to the ways of its community to the exclusion of others. Apart from the ways, the data he gets is also limited to that originated by the members of his cultural group. Later on when he encounters people from a different culture, he may feel overwhelmed and even threatened.
The book argues that if this pioneer team were more heterogeneous, the biases may get cancelled out. It is the equivalent of bringing up a child in a multicultural environment.
An AI, after it gets started, may, later in its 'life', encounter data that significantly deviates from the ones that it has trained on and also finds its established methods failing in handling this data. Depending on the kind of intelligence bred into it, it may exhibit behaviours that may be difficult to predict in advance and less than desirable. This wouldn't be the case if it was built up using an inclusive instructions and data.
Extrapolating the argument, it is important that humans too be exposed to different cultures as well as the common culture of a super-group like a nation during the early formative years, to avoid them displaying confused and roguish behaviour later in life.
Such bringing up will make a person more rounded without rough edges and spear points. Such a person will not have the desire to hurt or eliminate an alien behaviour on emotional grounds alone.
But yet, the fact remains that even this well rounded person and his consciousness is merely what he has learnt from his (more inclusive) society and gives no clue whatsoever about what it was like when it emerged into this world. Just as an AI cannot be described just on the basis of its hardware and OS. In fact it emerges only after initial instructions and exposure to the humongous data sets.
No comments:
Post a Comment