JimRaychrudhury
Ever since its inception, automated machine learning has caused lots of controversies. The ability of AI to develop new AI models leaves no one unbiased. On the one hand, there are those who are afraid of such cutting-edge developments, while the others are mainly focused on the benefits. Where is AutoML to take us someday? The possibilities are infinite and even unpredictable, but one thing is for sure: if Google has set their heart set on it, the sky is the limit.
To get a better picture of the abilities of automated machine learning, let us mention one example: Google’s contribution to chess AI, the famous AlphaZero. It takes a great deal of skill and, well, intelligence to be able to master the game. In the old days, computer programs were backed up by human knowledge, i.e. algorithms which provided computers with calculations needed to figure out the next best possible move. However, AlphaZero doesn’t need any human help in that way.
It was only provided with the rules of chess, but there wasn’t any input about strategies. In turn, AlphaZero developed its own algorithms and even defeated a world champion. Obviously, Google’s DeepMind AI Lab has done a wonderful job, showing us incredible abilities of artificial intelligence able to truly imitate our brain functions.
So, what’s in it for business?
First of all, AutoMl should enable us to save a lot of time when it comes to experiments and research. Start-ups using AI deal with pretty simple learning and finally get some value from the completed product. The whole process can be quite long compared to what is offered nowadays. AutoMI should already possess all the tools necessary, which speeds up the process and leaves more time for some other business targets such as product design, for example.
Consequently, companies will no longer have to put emphasis on their long research when presenting a product, which consumers aren’t that much concerned about anyway. Usually, what interests consumers the most is the final product, and they aren’t too keen to know the details about data collection and their interpretation. As visual beings, we are first attracted by the looks of something. We have just mentioned how a company will be able to invest more time in design, making the product more appealing. The right design is there to enable easier usage, and if everything goes right, it should be fairly easy to think about different options for the future.
Marketing may gain even more on significance, if possible. When the competition is fierce (in case a lot of companies will be using Google’s AutoML, and we see no reason why they won’t) and products are high-quality, the subtle marketing skills are there to the rescue. Sometimes it’s not about the quality, but the presentation.
Nevertheless, in order for automated machine learning to do its business, you have to supply it with the right information. AlphaZero chess program couldn’t have come up with all those glorious strategies if someone had failed to input the correct chess rules. For this reason, good research is essential in order to reap success.
Speaking of which, Google’s AutoML will probably make dreams come true for many an entrepreneur. No matter how complex it is and what amazing engineering is behind it, people are rarely impressed by it. However, what truly makes an impression is when the application is easily accessible. As strange as it may sound, AutoML is going to be just another useful application, though a pretty valued one for its advantages. After all, are we all in awe of our mobile phones? Do we see the decades of hard work, experiments and failures during the process of their invention when we look at them? Not really. The same goes for automated machine learning.
Conclusion
In short, Google is offering a great asset which has plenty of advantages to recommend it. The focus is now on input data, and the final product, whereas all the happenings in between will be fully entrusted to AutoML.
this is really good bro thanks.