Human Intelligence’s (HI) problem with delivering Deep Learning Experience (DLE) en masse precedes AI’s (Artificial Intelligence’s).
It’s a precedence that makes the development of strong or deep AI inherently problematic. Why? because it has no human model to automate, to follow: no structure, no guidelines, no human empirical framework to build its artificial legs or algorithms on.
The development of technologies that can deliver vast amounts of information/content quickly to diverse audiences did not just happen in a vacuum. Their development is and was innately connected to the human social condition. In fact, it reflects it.
Digital technologies are – in many ways – the reflective essence, the quantifiable empirical outcomes of the numerous unrecorded, unexplored, unheralded Deep Learning Experiences of their human developers.
Humans were and are at the epicentre of technological development. Yet more is known about the technology than about the learning experiences of its developers.
Explore the relationship between technology and its human developers – unpack the deep learning experiences that led developers, and still do, to their technological innovations – and the ‘traditional’ problem of developing Strong AI will be solved alongside the traditional HI problem of developing and delivering deep learning experiences en masse.
To put it another way, if digital technologies – from smart phones to Augmented Reality – represent the tip of an iceberg – the explicit evidence that the iceberg exists; the millions of hours of R&D that enabled the development of such technological innovations and shaped – maybe changed – the lives of those involved, lies hidden from view – immersed below the cacophony of digital delivery debates, marketing and advertising, and multi billion dollar industries.
Unpack the deep learning experiences that lead to, and CONNECT with, technological development – and/or any development – and pathways to developing and delivering Deep Learning – via AI or HI, will become clear very quickly.
The pre-requisites of any deep learning experience include:
- a learning experience that identifies (makes explicit) an implicit problem(s) for the learner – through
- an interactive exchange – that results in
- providing solutions (to the learner’s problem) – that
- ‘deeply’ connect to/with the learner.
For example, researchers seeking to develop AI capable of delivering Deep Learning Experiences maybe unaware of the relationships between their individual human social condition:
- their research problem, including methodologies, and
- humanity’s inability – historically – to develop DLE using Human Intelligence.
Awareness of such relationships – how they intertwine and reflect each other – needs to be DEEP. That is, awareness must lead to the ability to make CONNECTIONS – subjectively, structurally, practically to oneself, others, work methods and life experiences – in this instance, AI R&D.
Solutions that evolve out of a Deep Learning Experience (DLE) – evolve out of the CONNECTION between the learner and the (implicit) problem(s) they seek to solve. This connection lies at the heart of any DLE. Developing this connection – and fostering an awareness of it – is the role of the DLE collaborator.
A Deep Learning Experience (DLE) collaborator can potentially take any form – human or AI.
The key role of a collaborator in any DLE is to:
- isolate the learning problem
- focus the learner’s attention to/on the learning problem the learner needs to solve
- highlight connections between the learner and their problem.
- use these connections to support the learner to develop solutions to their problem(s).
Deep Learning is – from the onset – innately interactive and collaborative. It is multi faceted and multi contextual – it can, and should be able to, happen anywhere at anytime.
The key word in Deep Learning Experience (DLE) is ’experience.’ Why? because to learn deeply is to experience an understanding – and change – in ways that are relevant, personal, explicit, and make sense to the individual involved.
The individual context of a DLE is important, because what makes sense to us – and what feels relevant to us – might feel completely irrelevant and nonsensical to others. The interplay between individual experience and collective delivery is at the core of HI’s historical inability to deliver Deep Learning en masse, via any medium.
How can a learning experience – that seeks to change individual behaviour and raise self awareness of this change – be delivered individually and collectively? The answer lies in the method that is Deep Learning: CONNECT to the learner.
The connective tissue – for the individual to the collective, the local to the global – is the human social condition.
The human social condition has differences, but it also has a surprising number of connective similarities – albeit implicitly hidden in the social pathways of our everyday lives. Wherever those lives are lived – from war zones to luxury paradises.
The diversity offered by social pathways is a positive for developing DLE en masse – not a negative. It means the number of potential avenues for making connections to the diversity that is humanity – is immense.
An understanding and exploration of the human social condition can shed light upon why AI and HI find the development of resources that support Deep Learning problematic. It can link, align and find connections between:
- researchers, technology, end users, stakeholders
- the central problem (s) i.e., develop technology capable of delivering DLE; and
- it can offer solutions.
In short, solving the problem of developing AI capable of delivering deep learning experiences is an exercise in Deep Learning – for everyone involved. Deep Learning Experiences are simultaneously the method – the problem – the solution.
Dr. Louise Bricknell is an expert in developing and delivering Deep Learning Experiences via HI – to anyone anywhere. You can find her at louise@louisebricknell.com.au
Post script. I was looking for an image of an iceberg for this article, by happy coincidence I found one supplied by http://www.spindrift-racing.com/2015/jules-verne/drupal/en/log-book/icebergs-en. While the single handed Jules Verne trophy was never on my bucket list – the Volvo Round the World was – as I am writing this a friend of mine is getting ready for the 6100 NM Leg Six between Hong Kong and Auckland – go GIRL!