Software Solutions presents:
The Radical Revolution that is Deep Learning AI: Next Steps for Machine-Learning
December 18th, 2014.
Self-Teaching Algorithms - Deep Learning
Very shortly, machine-learning algorithms will no longer require human scaffolding in their learning. My prediction is circa 2018. This is important, becase it means deep learning, machine-learning by itself will take place without human intervention.
Expontential growth in the capabilities of machine-intelligence, think about it for a moment, artificial intelligence that self-teaches and self-replicates better models at how to be more accuracte at any given task. This usually means easily exceeding human performance in a very short time-frame, at a growing number of tasks that will become more and more complex.
- Deep learning is a set of algorithms in machine learning that attempt to model high-level abstractions in data by using model architectures composed of multiple non-linear transformations.
- Humans give the parameteres and AI can organize the data, like recognizing different kinds of pictures, writing captions for the pictures and doing a lot of things that we do in our everyday workplace.
To illustrate that I’m not simply talking pure fancy, I present to you this video:
By the date which you watch this video, deep learning AI will already have achieved well beyond those levels, that is how fast it is evolving. The field of deep learning is evolving in months, in tasks that would take a group of human beings years, if not decades, to accompilsh.
- For society this means, if you don’t have skills, your jobs could be in jeopardy in the next 15 years.
- Large portions of society will become automated with AI taking over a lot of customer service.
- Robots to accomplish a lot of tasks related to physical labor and uploading its day to an AI-bot net that can learn collectively.
- Huge revolutions in Medical and Educational fields of note.
- Even the data scientist may very well end up becoming obsolete.
- AI that mimics social behavior to provide social companionship. You may recall the 2013 film Her (Her, 2013 - Wiki). It became one of the highest-rated AI related movies of all-time (IMDB) and definately one of my favorite sci-films of the last few years.
From 2015 to 2025 we expect to see such a revolution in machine learning and automation of AI that it will become nearly unrecognizable to what it is today, and will appear to have “come out of nowhere”. This is because machine learning evolves along exponential lines, still quite unfathomable by mainstream audiences and even technology buffs.
What does this mean for the Future?
Because Deep Learning is all about learning multiple levels of representation and abstraction that help to make sense of data such as images, sound, and text. A point where algorithms can “integrate” the senses by itself , so to speak, is conceptually not far off. As AI requires less and less human cognitive input to self-teach, it becomes more autonomous and able to automate tasks in society, customer service jobs and changes the landscape of society from an inhert urban environment to a Internet of things “alive” urban landscape that is digitally automated.
For example, imagine sidewalks that self-clean and know when to warm to melt snow and imagine a thousand other things all automated and perfectly adapted to the changing variables of that environment.
When quantum computing and deep learning AI get “married”, expect the “singularity” to be near and AI to far exceed anything human intelligence is capable of. What does this mean for society? Well, we don’t really know. It’s unprecedented in human evolution. It’s way more radical than a simple invention of fire, tools or the industrial revoltuion, because it transforms everything in society. In effect, it makes cities come “alive”.
How will this entity, this hierachy of artificial intelligence unfold? How will it network with itself and other machine learning entities? This in effect, is all about the philosophy of the future. The problem is, the future is arriving so fast, inflormation flow accelerates at such a pace, that time begins to be distorted, or put another way, the future increasly becomes the present. With a radical shift to automation, how will unemployment levels adapt, how will captialism integrate AI, or more likely, how will capitalism be integrated into the AI reality. Sound like science fiction? In our lifetime expect to see a lot of things that would have been a water cooler joke 10 years previously.
“What Nice Eyes you have Mr. Robot”
Will AI become substitute Socialization?
In a world of Billions where a ruthlessly individualistic capitalism has been globalized, will alienated and vulnerable people have a means to socialize with cognitively aware AI? It’s highly likely, not only probable, but would be a huge innovation and consumer hit in the global marketplace. The educational value for children, the therapuetic value for isolated seniors, the potential is immense. So today I leave you with the trailot of the movie I was talking about:
Her (2013) Trailor