A.I. and Deep Learning: Breaking the Rules
Change is inevitable with the current state of tech disrupting every second of every day of our lives. We are not going back to the rotary telephone or the CRT TVs. The past, the present, and the future depends on how we want tech to impact and influence our lives as we live and breathe into our next stage of evolving. As a student at Holberton School, I had an opportunitiy to experience a day learning about Artificial Intelligence and Deep Learning.
Close friends Louis Monier, co-founder of Altavista and Gregory Renard, leading xBrain’s technical team, did an amazing presentation on giving an introduction into this world. While having personal interests into data analytics and business intelligence, I was pleasantly surprised at how big data is within the realm of Deep Learning. I thought the information getting everyone up to the current state going over the history of computers was engaging and not as abstract or boring as perceived to be.
Instead of telling you exactly what Deep Learning is, what I quickly learned at Holberton School is that it is a community driven space for self discovery to research the answers for yourself first a.k.a. Google. This is a great start for you to learn how to learn something you might not be familiar with. The presenters did give everyday examples of Deep Learning which includes Netflix, Spotify, YouTube recommendations, Google Photos by search, Siri for Apple Iphone, and the autonomous car.
We must realize that just a few years ago Deep Learning was not well known in the community or even a trend. There were two options that had a divide within the A.I. community on how to approach the better solution. The first option was to write a set of rules where it stays within those boundaries of their use, without the flexibility to change in a messy world, high costs and the inability to scale. The other option was for the A.I. to learn from the data by finding patterns automatically, ability to adapt, scale, and remain at low costs.
Some of the main points emphasized the two wanted everyone to understand for A.I. is that using a rules based approach is going to have limitations. A better approach would be to let the machine gather the data and learn from the data because “More data is always going to be better than a fancy algorithm” said Louis.
We ended the last hour with a fireside chat discussing the philosophical issues that arrive when applying A.I. and Deep Learning. The power of positive change with tech can easily swing to a negative if we do not take heed to our actions in society where we should always question our humanity. We saw great examples shown in Hollywood Sci-Fi movies of the future we dream of where robots do cater to our every need to the nightmares of killer robots. A solution I am an advocate for is education which includes exposure and awareness.
A big issue of job disruption was a topic heavily discussed that does not have a clear solution, but could change the typical concept of living as a whole. What would the world be like if robots disrupted the 9–5 workplace forever? These questions were very thought provoking and fascinating to explore to conclude that the future is unknown. We live in a time where there is so much data to be explored. The role of a programmer to problem solve is just a small piece of the grand scheme of things when it comes to caring for what kind of society we want to have together if we can create it. It will be interesting to see what the next ten years will be as this area in the industry of tech grows. Artificial Intelligence and Deep Learning might be called something else by then.