SBMartyn Richard Jones

Carmarthen 18th March 2017

This is a second collection of snippets; ideas and occurrences that are aimed to provoke critical thought and engagement.

As I stated in the first of these articles, “Most people who know me, at least in a professional capacity, will know that I am no unconditional supporter of Data Warehousing, never mind the more plague ridden swamps inhabited by Big Data, Data Lakes, Travelling IoT, Dodgy Data Science, Instant Data Scientists (just add water) and New-Age Machine Learning. So, many will not be surprised by this heterogeneous post-modern pastiche of ideas, impressions and bagatelles.”

I have extended the original scope of these posts to include any thought on themes of strategy, data, technology, process, business and humanity.

Please feel free to take me to task on any aspect of these reveries.

HERE’S A THOUGHT: How truthful and universally applicable is the statement that “When business solves a problem, it makes a profit “? Michael Porter seems to think it’s true. I have my doubts. What about you?

HERE’S A THOUGHT: Does anyone on social-media who is evangelising this ‘approach’ (of combining BI and AI) actually realise how extremely difficult it is to carry out any significant development in real AI never mind to bring about serious integration between AI and mainstream executive decision support? Then you have the problem of maintenance and extensibility? Tie that to commercial analytics/BI/EDW and see where that gets us… Ideas can be cheap, quick and unrealisable. Coherent, tangible and realisable strategies require some real critical thought… and the knowledge and experience to support it. What do you think?

HERE’S A THOUGHT: With mathematical Machine Learning making yet another comeback, how does one get over the perennial problem of greater numbers of false positives in highly-trained networks. I.e. In some cases the more data you throw at a problem of feature identification using ANNs the less ‘sure’ it becomes. Otherwise, if we can’t solve this problem, then won’t be simply revisiting the same problems real Data Mining has hit upon?

HERE’S A THOUGHT: Sometimes first impressions can be accurate. The uncritical Big Data, Machine Learning and IoT mania that swept the online world is yet more living proof of the validity of the observation that “you cannot argue people out of positions that they didn’t argue themselves into”. Big Data, especially for many people aspiring to work in that ‘field’, is more about belief in the vague, fuzzy and unverifiable, more than anything else. Ben Franklin was bang on the money: “Make yourselves sheep and the wolves will eat you”.

HERE’S A THOUGHT: I recently saw a video about how a Data Warehouse was shifted to Google’s BigQuery. But my thought was this… If you can migrate your Data Warehouse to Google BigQuery and with significant performance improvements, then maybe it wasn’t an Inmon Data Warehouse in the first place.

HERE’S A THOUGHT: A recent article criticised Hadoop’s failure to rise to meet people’s expectations. I think what is more accurate to state is that the absolutely scandalous Hadoop hype and the people who peddled it – some who continue to do so in the most shameless way, are the failures. Hadoopers may have been partly responsible for the surfeit of boloney, especially with claims regarding Hadoops ability to replace legacy Data Warehousing database technology. But, the worst offenders, by far, have been the indentured pimps and their astro-turfing of hype, half-truths, fake news and post-truth flim flam.

HERE’S A THOUGHT: Quite frequently, Business Intelligence initiatives fail because dashboards (or other artefacts) are not augmented with rational, coherent and credible narratives that actually explain what is being presented. BI requires an large degree of mature business competence, experience and knowledge in order to complement the more technical aspects of the solution and the data, and in order to be able to reconcile the need for data with an even greater need for executive-level explanation – e.g. “pretend you are explaining these figures and ratios to a highly socialised four year old”.

HERE’S A THOUGHT: To paraphrase the great Bob Hoffman, just when you think that if the Big Data babblers were to generate one more ounce of bull**** the entire f****** solar system would explode, what do they do? Exceed expectations.

HERE’S A THOUGHT: I recently had the misfortune to read an article on how the so called advances in AI and ML are supposedly reshaping healthcare. The reshaping of healthcare is a massive claim. Are there any limits that should be imposed on appallingly cynical and mendacious hype? Or should we simply give everything a free passage, no matter how atrocious?

Many thanks for reading and for allowing me the opportunity to shoe-horn the word ‘bagatelles’ into a tessellation of Big Data related curiosities.

You might like to consider joining The Big Data Contrarians