Yearnin’ for Machine Learnin’
Artificial intelligence. Machine learning. Bots. Agents. General AI. Cognitive computing. Big Data. Data network effects. Robots. Data moats. Neural networks. (dot) ai domain names.
The headlines, tweets, pitches, PR posts, and keynotes containing this verbiage are at an all-time high. It can be incredibly confusing, misleading, and overwhelming trying to understand what’s possible in 2017, why that’s impactful, and what may lie ahead.
More importantly, individuals want to understand how machines and data can change their day-to-day life (consumer experiences, routines) and career (job, organization, and industry).
Could this technology enhance an existing business process? Would it make us more efficient and save us money? Or could it enable a totally new product experience?
Is this technology going to take away my job? What changes will our industry have to make in the coming years?
What products do I use on a daily basis that rely on machine learning?
Machine learning is fundamentally a tool and a method, a tool to apply to a fundamental problem or a new method available for creating a differentiated product experience. Working backward from the customer want or market need, it should simply become a well understood capability that is evaluated along with all other options (another technology, changing nothing, not right now, etc). At the moment, that’s not the case as it is inserted far too high up in a decision making calculus.*
In the end, as this technology gets more approachable and robust, the number of companies using machine learning in some aspect of their business will far exceed the number of ‘machine learning companies’. Developing and releasing a mobile application didn’t make you a ‘mobile’ company, it was either a better way to solve an existing problem (hailing a ride) or enabling a new set of product experiences (digitally send money to a friend). If it isn’t either of those, users and customers won’t (and shouldn’t) care for long.
I see this as like ‘mobile’. It’s not an industry, it’s table stakes. Every startup should have an opinion in ML/AI.
I look a lot for companies that apply ML, or could one day have a data set that would use it.
Two years later and I feel both better equipped to start thinking aloud about this subject and even hungrier to continue learning about its impact on a technological, product, and business level.
Technology: what is the state of the art? do I need a state of the art implementation? what is the cost (time/energy/money) of implementing this ML system? do we have this historical data available to train on? what teams will build + manage this tech? what additional tooling is needed? is it possible to deploy this model to our mobile clients? how do we respect user privacy? how will we label our data?
Product: will this publicly affect the user experience or is this an internal process and workflow enhancement? what additional data should we be collecting? should we A/B test the new feature on certain cohorts? what is our fallback if the model fails in production? do we allow for users to provide feedback in our interface? how do we explain a model’s behavior?
Business: what does the product marketing effort look like for this? have we trained customer support on new potential errors? what is the ‘ROI’ of this new implementation…is it even possible to quantify? does this help us meaningfully differentiate from competitor A and B? is this a build or buy decision? does it benefit other ML initiatives in the future?
Before one can start to sip from the firehose of news, product releases, acquisitions, integrations, ‘partnerships’, hardware innovations, company launches, conference announcements, and research breakthroughs, a basic collection of mental frameworks have to be in place.
The intention is for this to become a set of living resources primarily focused on enabling product and business minded folks to effectively navigate, learn, and build in this new landscape.
It’s time to make these headlines a bit more real and actionable.
v0.1 Table of Contents:
Updated weekly. Notes, curation, and writing.
The Feed (ongoing daily notes)
The machine learning, learning landscape:
Impact on organizations and industry talent:
Machine learning primer for executives:
Deep learning present in today’s products and technologies:
Machine learning metaphors:
Common questions people have regarding machine learning:
Proprietary data set collection techniques:
Machine learning deployed in production:
A list of open machine learning questions:
Ethics in AI:
Interesting early stage projects:
A resource of resources:
Machine learning newsletters:
Who to follow on machine learning twitter:
#SuperFrontier:
*Like it or not, association with artificial intelligence and machine learning will continue to be a tactic for companies of all sizes in effort to build brand equity, generate press opportunities, attract talent, and keep up with the perceived competitive landscape.
**Machine Learning as a Service (MLaaS), infrastructure, and developer tooling in the technology stack portion of this image. HT Shivon Zilis
