I challenge the viewpoint that AI-based home robots will have a promising future. 
In the home AI robot session in “Stanford One Hundred Year Study on Artificial Intelligence”, it was said that AI robots will flourish in the next couple of years as a result of the improving technology; however, most of the points made are based on the hardware improvement, such as more powerful CPUs, more efficient 3D sensors, and cheaper robotics arms. In my opinion, hardware development will not lead to the success of home AI; it’s the software that matters. 

To figure out how to create successful home AI, we need to first understand why we want to have an AI assistant. Personally, I want to have an AI assistant that will speed up my living efficiency, but usually, it simply slows me down.
I believe most of us have had the experience of talking to an AI assistant and hope to hear back the information we need. In advertisements and YouTube videos, AI assistants behave like a real human; based on our input, they can answer our questions with exact information we need. If that is true in reality, I would certainly use them for every job I am doing, but that fact is, I almost never used them, except for those few days I had in my freshman year. 

In the first semester of my freshman year, there was a time when I frequently used “Google Now” to look up weather. On each morning I got up, I would turn on my phone and asked it: “OK, Google. What is the weather today?” This was cool at the beginning, because the magical voice answered from my phone. “Here is the weather today”, and the resulted information displayed on the screen both made me feel like a king. Nonetheless, days later, after the freshness disappeared, I realized how stupid it was. First of all, in order to start AI assistant, I had to first press the search bar on the top of the screen (it is actually faster than saying “OK Google”) and after it was launched, I had to get close to my phone’s microphone, and slowly speak “What is the weather today?”. Next, I would wait a few seconds for the Google server to process and return information, which depended a lot on the network quality. Then, a list of information would show and finally I saw the weather. The inefficiency of this whole procedure drove me mad, especially when I was catching a class in the morning. I immediately quit using “Google Now” and installed a weather app instead. Now, whenever I want to see the weather, all I need is just to unlock my phone. 

From my perspective, the current software support for AI assistants is far from enough and unfortunately, AI assistant appears to be the core function of nowadays home AI robots. In most of the time, AI assistants are only doing a google search for the message you put in, which is not very productive. There is no doubt that AI assistants will be useful under certain circumstances, such as when people are cooking and unable to type, but they are too constrained to make the whole industry flourish. The hardware development will not be enough for the advancement of home AI industry nowadays and better ways of operating AI assistants should be developed. Although I believe home AI robots will be indispensable for future human beings but based on the points given by the article, I would not see it happen in a near future. 