Should We Worry About Bias in AI?

Arrangement of human head, key symbol and fractal design elements on the subject of encryption, security, digital communications, science and technology
9 Predictions About AI That Might Surprise You
December 12, 2016
OLYMPUS DIGITAL CAMERA
With the Right Investments Made, Intel’s AI-Future is Looking Bright
December 13, 2016

Should We Worry About Bias in AI?

adobestock_58063500

The Merkle have written a very insightful and important article about AI and how bias will affect it in the same way that humans do, and that this is something we need to address. However, this does not mean that AI has a less bright future ahead.

The Merkle writes that unlike what most people might assume, AI can be affected by bias, an aspect that is necessary to concern when thinking about this new technology.

“Even though the concept of artificial intelligence is still in its infancy stages, experts are already warning about the possible pitfalls”, the article reads. It emphasises that it is good that people are not hesitant to address potential flaws in this technology that is now more hyped than anything.

“These drawbacks, however, do not mean that AI has no bright future ahead”, the article reads.

The reason why it is important to address these bias concerns is because they can change they way we think of AI in general. With most cases of applying AI, there is a need for massive amounts of data, and if this data is already biased, the result will reflect this too. It will make the technology automatically assume that what it is reading is correct, which will contribute to further bias.

The Merkle writes that even though AI is capable of learning independently, some hand-holding will be involved in the early stages of deployment, which is necessary to avoid bias in any way.

“Developing an intelligent cognitive system is one thing, but making it objective is rather difficult. As people are involved in the development of these new tools, there will undoubtedly be some form of bias involved”.

We still don’t know whether AI will act as its creators, but it is certainly a concern that should be addressed as developers will have particular preferences, mannerisms, and manners of communication.

The future of AI is evidently looking bright, the article reads, but it does require a run-through of several iterations before it can be looked at as a completely separate entity. However, this is not necessarily a negative thing, because letting these creations free without any restrictions could have catastrophic effects.

“Bias will be bias, an area in which humans and AI are not all that different”.

For the latest news and conversations about AI in business, follow us on Twitter, join our community on LinkedIn and like us on Facebook

 

 

Leave a Reply

Translate »

We use cookies. By browsing our site you agree to our use of cookies.Accept