2018 has already been an explosive year for new technologies like the Blockchain. With that in mind, it’s easy to wonder when AI will undergo a similar period of rapid growth.
According to a Forbes article published yesterday, next year will be that year. Through examining their specific claims on the subject, it may become easier to see why this is at least somewhat likely. Overall, it’s important to start with the current boom in GPU processing power that is already pushing Blockchain projects to great heights. This, in connection with the expected advent of widespread Quantum computing, means that in the near future, we should have at least double the computing power that we currently possess.
Now that you have this in your mind, it will likely be significantly easier to understand the four main predictions that Forbes shared yesterday.
Firstly, in referencing a venture capital fund called Emergence, they state that in 2019, business communications will shift away from those that are more voice dependent and towards video-based platforms that are augmented by different sorts of AI assistants.
A good example of this might be a Skype-like platform that allows businesses to understand the body language of the person on the other hand, with the help of an AI system. If this sounds like some sort of terrifying big brother-like situation to you, then you are assuredly not alone in thinking so. In summing this prediction up, Emergence’s conclusion is fairly logical. When we all know that AIs can analyze us based on our physical actions during a meeting, then why not have all meetings in-person?
In the article’s second prediction, however, a contrasting idea is brought to light.
At this point, Forbes actually mentions a quote from Zoom, which is a well-known competitor to Skype in the video conferencing space. Essentially, what they believe about AI in 2019 is that due to the development of what they call smart features, video conferencing will continue to be the norm for business meetings, if not more so. By smart features, it appears that Zoom is referring to anything which AIs add to a video conferencing platform to give it some sort of improved capability. In specifying what some of these might be, Zoom apparently mentioned the skill of AI in transcribing voice-based meetings to text. If you have read any recent articles about AIs and transcription, then you may already know why their skill in this area is significant. Recently, several studies seem to have proven the ability of AIs to transcribe as well as humans. In the interest of being realistic, however, it is necessary to add to this by saying that this is not exactly the case. While AIs have reportedly gotten very close to humans in terms of their transcription skills, this is really only true in tightly controlled situations.
In addition to referencing this particular situation, Zoom also mentions how AI-driven facial recognition should affect video conferencing. With this, they essentially seem to claim that in less than a year, AIs will be able to remember the habits of certain people who use specific conference rooms in order to help organizations to better plan these events. How this specifically might be possible on a technical level is not explicitly mentioned, though it is reasonable to assume that it involves a bit of work in teaching AIs to better recognize emotions.
Following this discussion, the writer mentions what is perhaps the most logical prediction of them all. With the growth of AI, the need for positions like Chief Information Officer and Chief Data Officer will grow as well. Accepting this as at least being close to a fact is fairly easy to do. As Artificial Intelligence systems gain more of a foothold in terms of how we do business, we will need more people with a deep understanding of what makes them tick. Inside of this proclamation, which Forbes sourced from an executive at McAfee, it is also stated that the usage of supervised learning will grow at the same time.
Therefore, if this third prediction is to be believed, in 2019, we will not likely be anywhere close to a reliance on unsupervised learning principles and as a result, reaching AGI. After this, the article swiftly moves to its’ final prediction which generally states that in 2019, we will begin to accept the idea of AIs being doctors one day. In describing why this will be so, Forbes mentions a quote from the Chief Product Officer of Looker, which is a fairly well-known data analytics firm. The problem with this quote, however, is that it assumes that AIs have already surpassed humans in this field. If that were the case, then we would reasonably already be in the Age of General Intelligence. Therefore, with all of this in mind, we offer one prediction for you for AI in 2019.
The overall AI space will continue to improve but as it does, it will become more and more important to reign in rampant speculation regarding AI’s evolving use cases. In connection with this, we still have not reached the point at which AIs can be said to be truly intelligent. If our reasoning here is not exactly clear, think about what makes humans intelligent. In the most general sense, we are because we can create our own paths towards solving problems without consistent guidance from others. Furthermore, the solutions that we come to at the end of these paths do not have to be pre-determined in any sense.
