Myths surrounding artificial intelligence. There’s a lot of them. Here are 5.
1. Artificial Intelligence is going to shrink the job market
This is a myth that is commonplace in any conversation featuring artificial intelligence. The technological revolution that we are currently undergoing which includes but is not limited to artificial intelligence is, to simply put it, a near identical replication of the Industrial Revolution. There is a term used in the world of business and economics known as ‘creative destruction’ and it refers to the phasing out of old technologies and replacing them with new. This inevitably will mean the phasing out of old jobs and the introduction of new. A common example is that of Henry Fords automobile production line. This new method of car manufacturing revolutionised production and factory methods and invalidated countless jobs, whilst creating a plethora of new ones. Innovation is an evolutionary process that seeks to remove inefficiency, and artificial intelligence is no exception. The overall result will be a changing job market. Not a shrinking one.
2. Artificial intelligence is objective
Artificial intelligence will only ever be as objective and unbiased as the data and processes in place to train it. Any process that requires human interaction or an element which has been subjected to human interaction will result in a bias. That includes the building and training of an artificial intelligence entity. The data that AI is trained on will inevitably host a bias that shows through in the technology and so to say that AI is entirely objective is an entirely false statement. The best that can be done from our end is to limit the bias that is present within the data that AI is trained on and to be `self-aware of any bias we may be feeding our AI.
There are plenty of examples of where societal bias has shown through within AI. A recent example is that of COMPAS, a predictive algorithm used in the US court system to calculate the likelihood of a defendant’s recidivism (reoffending). It is a risk assessment algorithm. However, it was unfortunately discovered after a few years in use, that the algorithm was inherently biased against black offenders, and gave preference to white offenders. 45% of predictions for black offenders were false positive whilst only 20% of predictions for white offenders were false positive. This simply was just a bias within the model and highlights how harmful this myth really is. Artificial Intelligence as it stands now is subjective.
3. AI is not useful for my business
AI is absolutely not just limited to the manufacturing industry. No matter what business area you are in, there will be some way in which artificial intelligence can assist you and excel your company. If your business has any form customer support, then there is a place for AI in the form of automated chatbots. If your business has any sort of online presence, then there is a place for AI to maximise you search engine optimisation (SEO). The best thing is that these technologies aren’t difficult to integrate within your online presence. At AI Commerce we integrate artificial intelligence with e-commerce stores who use Shopify. You can learn more about how we do this here. The bottom line is that every business should be looking at how artificial intelligence can play a part in its strategy. Not doing this will inevitably place your company at a competitive disadvantage.
4. AI and Machine Learning are the same thing
The terms artificial intelligence and machine learning are often used interchangeably however they are not the same. Machine learning is one of the main subfields of Artificial intelligence, and so referring to artificial intelligence doesn’t always necessarily mean one is referring to machine learning. Artificial intelligence is a very vague concept with blurred boundaries, whereas machine learning refers specifically to the sector by which algorithms are designed to adapt and ‘learn’ through the use of large datasets and past experience without any formal instructions from engineers. Other sub-fields within artificial intelligence include computer vision and natural language processing.
5. AI is an autonomous technology
To say that artificial intelligence is autonomous would be to say that artificial intelligence does not require human intervention to be maintained or applied to a given problem. As the field stands today, we have not yet reached this stage. Machine learning models still require engineers to feed the data which they are trained on, architect the problems upon which they are applied, and then adjust the model based upon its results to remove bias etc. Whilst machine learning models are very impressive, they require human input to improve and evolve, and so therefore cannot be classed as autonomous technology. Autonomous artificial intelligence often refers to the idea of artificial general intelligence which is the science fiction artificial intelligence that can comprehend and learn as a human can. Whilst this may be achievable in the distant future, we are far from this being a reality today.
So, there you have it! 5 common misconceptions about artificial intelligence.
The field of AI is often stigmatised and can receive a lot of scepticism which hinders innovation and the adoption of useful and productive technologies. By educating others about this ill-informed stigma, artificial intelligence may be seen for the amazing technological tool that it really is. Here at AI Commerce, we use cutting-edge artificial intelligence technologies to make Shopify stores sell more. If you want to learn more about our services, you can find us here.