A Personal (and Professional) AI Journey
- A Personal (and Professional) AI Journey
We are moving from ego-systems (profit driven) towards ecosystems (purpose driven). The bigger the organisation the bigger the challenge…
AI will push humanity towards their purpose, more and more people will follow their heart as a result. And smart technology will help human beings to develop themselves in that direction.
Entrepreneurs live in a relatively unstructured environment.
We wake up each day and are presented with perennial challenges around how to create and then capture ‘value’.
If you are unable to create value (in the eyes of users / customers), your idea is unlikely to be around for very long.
If you manage to create value, but fail to capture a sufficient proportion, the outcome is going to be similar (unless you have established a Wikipedia-style governance structure where nominal value may not be required).
This problem is particularly acute where you are either:
- Trying to create a new market from scratch
- Attempting to disrupt an existing market
(or even worse – both).
So in that context it’s not often an entrepreneur comes across a lever that is so big it can’t be ignored, and above the lever is a huge sign that says “Pull me”.
Alongside the lever are a set of instructions, which make it clear that pulling this lever will remove much of the ambiguity you are facing, providing a structure to your day-to-day activities.
The instructions are in effect a ‘micro-economic roadmap’ that can be shared with stakeholders so that everyone can grasp not just the destination (the vision you have been describing), but also the journey.
For myself artificial intelligence (AI) is that lever, and I have decided to begin writing a regular blog about the personal (and professional) journey that myself and the whole Mclowd Team are embarking upon.
As Christian Kromme points out, ‘AI will push humanity towards their purpose’.
In the process it will enable challenges (economic, ecological or sociological) to be overcome with a fraction of the resources that would otherwise be required, and in a fraction of the time previously allotted.
The goal of this series of blogs is to empower others by illustrating that someone like myself (with little if any technical skills in computing) can leverage AI in order to achieve goals that might otherwise seem improbable.
This first blog will focus on describing some of the key concepts that underpin an understanding of AI.
In subsequent posts I will begin to describe the way myself and the Mclowd Team will use AI to achieve the goals of the Mclowd Community.
After an initial period of promise (dating back to WWII) AI entered a long period of stagnation.
It has only been in the last 10-15 years that increases in computing power (and flexible access arrangements via platforms such as Amazon Web Services and Google Cloud) have powered rapid innovation.
At the same time, the cost of deploying AI has collapsed towards marginal cost (which in most cases will be zero).
This journey can be summarised by the following diagram:
The above diagram also helps to clarify the scope of this ‘AI blog’, which will focus on machine learning, as it is the scope that I am (for now) most comfortable with.
What is Machine Learning
The best definition I have found is on the Elite Data Science website:
“Machine learning is the practice of teaching computers how to learn patterns from data, often for making decisions or predictions.”
This definition also helps to understand the renaissance of AI over the last 20 years, which is a consequence of two base elements being fused:
- An enormous amount of raw data
- Virtually unlimited computing power
The emergence of this foundation has not surprisingly corresponded with growth in the data sciences, and I have sought to illustrate these relationships in a simple heirachy.
The insight which machine learning can extract from data spans a number of fields including:
- Image / pattern recognition
- Measuring / estimating the probability of future events
- Human to machine communication (via natural language processing)
Examples of Machine Learning
But perhaps the best way to illustrate machine learning is with real world examples, and as I have come to realise, machine learning is a lot more common than most people think.
Apple’s voice assistant Siri uses machine learning algorithms to predict that what you said was “call Jack” (rather than “call Jane”) because it has heard the phrase ‘call Jack’ enough times in enough accents (along with your phone contacts) to be confident as to your intent.
Extending the mobile analogy, the iPhone can read your fingerprint and acquire a level of confidence based on the pre-existing data it has stored.
Of course Apple is now doing away with the home button on its iPhone in preference for facial recognition, but the principles (as to probability / functionality / usability) are exactly the same.
Image recognition is also being used heavily in online marketplaces (such as www.carsales.com.au) to flag user content for review and improve search results / user experiences.
The Australian Centre for Invasive Species Solutions is using machine learning to build intelligent devices which can see, think and act in order to target specific pest animals.
In the Australian context the eradication of pest species such as foxes, rabbits and feral cats is – in my view – a perfect example of the huge upside which this technology offers.
This first blog in the Mclowd AI series’ was designed to lay the foundation for a collective journey, such that readers acquire a basic understanding of the key concepts which underpin the technology and its application.
The next episode will revolve around an internal session that is being held as part of the Mclowd October Team Meeting.
The goal of that session is to identify an MVP (Minimum Viable Product) that can be deployed into the existing software to automate the classification / processing of bank transactions.
Beyond that the sky is the limit for the deployment of AI within the Mclowd Community.