Are we ready for self-driving cars?

Cars

Self-driving cars are becoming real and various predictions have them being mainstream over the next decade.  But it is interesting as to if we are ready for this innovation in transportation and how we will react when accidents involving self-driving cars invariably happen.

I am not a physiologist, but the human element of technology adoption is of course fundamental.  And a potential issue for self-driving cars is that, of course, roads are quite dangerous.  We hear about people being killed at the time.  In fact based on 2013 figures 0.02% of the world’s population (1.25 million people) are killed on the roads every year, over 3000 people a day globally.  Yet this is a risk that we accept and strap our most loved ones into our vehicles for journeys as trivial as going to the beach or getting ice-cream. 

So how do we accept and process this risk? As a species we seem to be able to deal with risk by applying a contrived analysis which results in use determining that bad things will never happen to me.  We hear about accidents, but we believe we are each a better driver than those involved, we are move observant, we have faster reaction times, there are lots of dangerous people on the road any my job as a good driver is just to avoid them.

I have no doubt that self-driving cars will make the roads safer, probably significantly so.  But accidents will still happen, it is improbable to think otherwise.  The software powering self-driving cars is typically prediction based – prediction = probability = (small/tiny) potential for being wrong.  How do we process the risk once I no longer have the advantage of my "better driving" – when everyone’s tech is the same as my tech and the responsibility for a safe journey is out of my hands?  For some, perhaps many, this will make the risk a lot more paramount and the act of going out for ice-cream somewhat more concerning.

The Maturing Field of Big Data

Reddoor

I remember when I was 19, I was working for an electricity utility as a DBA.  I was putting in lots of hours and partly as a reward, and partly because they didn't know what to do with me, then sent me off to a knowledge management conference.  Well, when I got back I was an “expert” in knowledge management and it was going to change the world forever.  I convinced my boss to get me in front of the executive team during their next board meeting.  I was on fire, delivering what could only be considered a stellar presentation on knowledge management.  At the end of the presentation I looked around the room expecting excited faces and many baited breath questions. Instead, you could have heard a pin drop.  They were staring at me not knowing what to say, until one of the executives jokingly asked if I had learnt why Microsoft Word keeps crashing while I was at the conference.  Then they moved on with their meeting, and knowledge management was never discussed again during my tenure.

For a long time after I was thinking how foolish they were, to ignore the technology which was going to change their business. I was literally handing the insight to them on a plate.  However, over time as I got more experienced my view started to change.  At some point many years after I had left I realised I was trying to sell them a solution for a business problem they didn’t have. 

As a wide eyed techie I had an assumption of perceived value, they had thousands of user files on network servers, knowledge management allowed you to structure, access and understand these files in a better way and to me that sounded like it had immense business value, although I wasn't exactly sure what this business value was and couldn't articulate it any deeper than “insight” or “understanding”. And because of this I had failed to put this into any context they cared about, how KM was going to sell more electricity or prevent outages.

Fast forward to present day and I have seen a similar repeat of my experiences across the industry in relation to Big Data, and certainly some of the commentaries on Plantir resonate.  Big Data has been primarily IT lead on the assumption that if you get enough data into one spot there is significant inherent value in it.  You can find lots of web articles about this, about finding “needles in haystacks”, about discovering previously unrealised relationships, about “monetizing” data, about understanding customers better.  But when you try and dig into the detail of what this actually is, there is certainly less information available.

Thankfully, Big Data is starting to move out of the hype phase with the related, but separate fields of Machine Learning and AI starting to take over as the topics generating internet buzz. 

Gartner_Hype_Cycle

But the hype is always a good thing, for a period of time, as out of it we now have awareness, technology, toolsets and capability to develop business solutions using Big Data.  But as the field has matured we also need a mature view to be successful. This includes:

  • Big Data must be business led rather than IT led.  They must be attempting to solve problems that the business cares about and has meaningful impact.  IT is an important part of the solution but not the driver.
  • Solutions must identify value that is not easily identifiable using simpler/less costly methods.  For example, say you have a factory that makes red doors.  Sales have been great but over the last few months sales are declining.  To solve this you using Big Data to identify that customers are growing tired of red doors and now they have a preference to buying blue doors.  Did you really need Big Data to solve this problem?  Do you think if you spoke to your #1 door salesperson they wouldn't be able to give you the exact same information?
  • The Big Data solutions must lead to actions that the business can undertake.  If they have a red door making plant, perhaps they can modify to make blue doors.  But they might struggle to start making cheese.  Big Data has to provide insight within business context.

This doesn't mean that Big Data is becoming boring, far from it in fact.  Instead this maturing means we are more focused on delivering data driven solutions that are going to have a real impact on the world around us.  For any analyst/data scientist that has to be more exciting than simply churning data for data’s sake.

SQL Server on Linux is not an Oracle killer – but it is a great buffet dish

Buffet2

The planned SQL Server for Linux product was been announced a few months back.  While this isn't due until 2017, I have read numerous commentaries and thoughts on why Microsoft is doing this.  Several of these are suggesting this somehow will cause a mass migration of Oracle DB customers to SQL Server.  This prediction is silly and shows a lack of understanding about how enterprises select their database platforms.  SQL Server is a great product but Oracle DB is also a great product. I have no doubt that the majority of customers running it today either want to be running it or need to be running it as it supports the applications they are running, providing the operational objectives they have at a cost they can justify.  Just in the same way that Microsoft SQL Server customers do.

So what is SQL Server on Linux about then?  To me, it is more about breaking the perception that SQL Server should only be considered as part of a solution if you are using a full Microsoft stack.  One of the problems with enterprise software has been, in the past, heavily orientated towards buying into that vendor’s vision.  It has been analogous to dining on a set menu at a fancy restaurant.  You had the ability to choose the restaurant, but the courses are largely pre-determined by the restaurant. 

This reminds me of once when I was in Lyon (France).  Lyon is well known for its culinary excellence and as such I invited a group of industry colleagues out for dinner at a Michelin three star restaurant to enjoy the best of Lyon had to offer (actually it was me, a data entrepreneur, a senior exec of Greenplum and a Harvard professor).  When we arrived we found there was no menu, we would be served what the restaurant served.  As it turned out, one of my guests was vegetarian which was met with some disgust by the staff and he was told, in a fabulously French way, that he would have to wait until the cheese course to eat!

Anyway the point of the analogy was that, we chose the restaurant, we bought into the stack and each layer/course was what than vendor decided to provide.  While it met most of our needs one of our party wanting to do something different wasn't well accommodated – and I can give you assurances he would not be returning to that restaurant! 

One of the way the cloud changes things is that customers have much more freedom to combine the most suitable offerings from across the IT landscape into customised solutions within a common environment. At the risk of extending the analogy to a ridiculous level, the cloud can been seen as a technology buffet.  Cloud vendors are presenting their offerings and offerings from others, but the customer is free to select and pair whatever they want with whatever else they want.  You want bacon and foie gras, no problem!  You just want scalability with a massive plate of fries, go for it!  You want SQL Server feeding into Spark, enjoy.  The cloud is about freedom of choice.  Vendors wishing to maintain relevance need to be embracing the freedom of choice.  Vendors are starting to realise they may not win at every layer, but having independent options prevents them automatically loosing at every layer.

To me, SQL Server on Linux is Microsoft breaking the perception of a stack dependency and making the use of SQL Server a much more independent decision.  SQL Server is a leading edge database product and the SQL Server on Linux release gives it the ability to win in solutions where other parts of the Microsoft stack, including Windows don’t.  And importantly, it is just further evidence that Microsoft is backing its strategy to win in the cloud by maintaining relevance.

Is the Microsoft HoloLens the next big thing in Analytics?

If you haven’t seen it yet, you should check out the Microsoft HoloLens demos. While it is not widely available yet the developer edition is out and Microsoft is working with their partners to get applications built that make use of the holographic and augmented reality potential.

At first the HoloLens may look like an expensive toy, designed for gamers. Or you may see it as a tool limited to designers. But moving past that, the Microsoft HoloLens has significant potential in the field of data analytics. One of the key challenges of Big Data has been turning the outcome of analytics into a humanly digestible format, so it can be easily explored and understood. However there is a limitation on what you can show in 2D within the confines of a computer screen. The Hololens has an opportunity to change this. Adding an extra dimension to data visualisation combined with a 360 degree view may fundamentally change the way we present data in the future. In addition to data exploration, augmented reality may allow the outcome of analytics to be attached to the real world objects they relate too.

This is of course somewhat dependant on if Microsoft has got the Hololens right and don’t follow the same tease and revert path that Google famously did with the Google Glasses. If the HoloLens really is ready, this is a space that Microsoft can own from the get go with first mover advantage.

Blockchain, Blockchain, Blockchain, Oi, Oi, Oi!

Sydney_bridge

One of the fastest moving technologies of 2016 is Blockchain.  Put simply, Blockchain is decentralised trust system for ensuring transaction validity without a central authority.  The use-cases for Blockchain are far reaching as it is essentially a data platform on which any applications that require trust for the “exchange” of information between multiple parties can be built.  And it has just been revealed/confirmed that it has Australian origins!

The Blockchain methodology was designed to underpin Bitcoin but has since gained momentum in its own right, in fact arguably Blockchain is a more important innovation than Bitcoin itself.  Certainly banking and FSI in general are all over Blockchain, with CBA and 10 of the world’s largest banks simulating trading using Blockchain, and the ASX working on a Blockchain based system for Australian equities.  But these are just a couple of examples. The interest, momentum and pace surrounding Blockchain is quite astounding.

Blockchain is still in its early days and doesn’t have all the issues solved yet.  Scalability can be a challenge but these problems will be resolved as Blockchain technology is evolved.  In terms of resource, there is a lot of technical detail on the web about how Blockchain systems work, however if you’re looking for resources on the “why” for Blockchain a couple of good ones include:

If you’re involved in building or operating trust based systems that exchange information between multiple parties you absolutely need to get up to speed with Blockchain.

Will Automation take my Job? Well, Maybe….

Will Automation take my Job?

Automation is a business transformation technology that involves innovations in the field itself, but more recently leveraging innovations in the areas of AI, Machine Learning and Big Data. And as all of these fields gain maturity, pundits are naturally playing forward the impact and making predictions about job losses across various industries directly as a result of automation.

Reiterating the title of this post, “will automation take my job” I think the answer is a clear “maybe”. But job loss isn’t the only outcome of automation. My experience has shown that many organisations are seeking to increase the value of the output of their internal workings, and often key employees are constrained with low value tasks. In IT this is particularly true, where many employers are seeking proactive innovation and thought leadership from employees in their respective areas. But often this is not being realised as they are consumed with lower skill, high occurrence tasks that are important – but are not producing an ROI to the business. IT is just one example, the same problem can cross many industries and skill sets.

Automation of Today

Today, automation can be good at undertaking pre-planned actions when pre-defined conditions occur. Which means certain types of roles, that are formulative in nature, lend themselves to automation. But trying to improve the efficiency of these roles is not necessarily new. Many organisations have already spent effort reducing the associated costs, sometimes replacing higher cost resources with lower cost alternatives. This transition typically required organisations to document the process aspects of these roles in detail, naturally this feeds well into the foundations of an automation drive. And this is not necessarily limited to the lower end of the pay scale, I am sure there are a number of people in high paying roles in FSI, trading, banking etc. that are beginning to see components of their role replaced by automation.

Automation of Tomorrow

Looking forward, automation is beginning to become more adaptive and use machine learning and AI more broadly to make judgement calls. Bots may understand typed and spoken language as input. Routines may use analytics and prediction to select the best cause of action to a specific situation. This broadens the scope of the application of automation from tasks, which have clear black/white outcomes to those with shades of grey requiring intuition calls.

"If you are doing the job of a robot today, then it is logical to think that computers may one day replace you. But the question is, do you want to be doing the job of a robot to begin with?"

So is this all doom and gloom? I think this is definitely an approaching wave of change that is going to impact on areas of the workforce. Over time this will phase out some roles, and aspects of others, but it will also result in creation of new roles and the improvement of others. Contrary to how if can sometimes seem, most organisations are not just trying to cut costs. They are instead usually focused on ensuring value is being created for both their customers and their shareholders, and driving their competitive advantage. While this does mean reducing costs where practical, it also means making investment in areas that continue to drive growth. This should therefore also mean new jobs, new opportunity and more innovation across the board.

What to do?

But it does mean change is likely for some, and change can be very unpleasant. To ensure you are ready for change I think you need to take an honest look at your current role to determine if it fits the model of a function that overtime could be automated. If so take the opportunity to begin preparing for the change, developing skills and experiences that will ultimately be of higher value if/when organisations begin to adopt automation as a means to increasing value.

Is Amazon about to Disrupt the Database Market?

My LinkedIn feed shows me Shawn Bice, former GM of database systems (SQL Server) at Microsoft is joining Amazon AWS as VP of Analytics. Assuming accuracy, Shawn has joined the likes of Hal Berenson (on the former Microsoft luminaries behind SQL Server), Raju Gulabani and Sundar Raghavan.

While AWS has for many years provided support for common database platforms via their EC2 and RDS options, more recently they have released their own transactional database platform AWS Aruora, and the AWS RedShift data warehousing platform. And to get you there, they have also recently released their database migration service for on-mass on-premise to cloud migration.

AWS seem to have realised that a keystone in winning in the cloud is winning the database. In the data centric world ahead, the data platforms are going to become core to how applications are architected and ultimately deployed. Within the cloud providing a comprehensive set of data services with (semi-) seamless integration, rapid deployment and op-tap scalability will be compelling in convincing developers and organisations to “buy into” that vendors stack.

AWS are actively hiring some of the best and brightest in database for what could be a double whammy if they can get it right. The last time I looked I think the database market on it’s own was a $30b+ market, but in the cloud winning with the database also likely means winning a customers complete cloud stack.

Of course, Microsoft and Oracle are formidable opposition and are arguably ahead of the in terms of developer and enterprise buy in. So it is not necessarily and easy path for ahead.  

I think I have been saying this continuously for the last 15 years; but it is (still) an interesting time to be in database.

What is the biggest challenge for Big Data? (5 years on)

Five years doesn't half fly when you’re having fun!  In this post from 2011 I highlighted some of the challenges facing the “big data revolution” centring on a lack of people with the right skills to deliver value on the proposition.  Fast forward to 2016 and this not only remains true, but is likely the key issue holding back the adopting of advanced analytics in many organisations.

 While there has been an influx of “Data Scientist” titles across the industry, generally organisations are still adopting a technology driven approach driven by IT.  The conversations are still very focused on the how rather than the why, it is still all very v1.0.  There is still a lack of the knowledge required to turn potential into value, value that directly affects an organisations bottom line.

This will start to sort itself out as the field matures and those who understand the business side of the coin become fluent with big data concepts, to the point they can direct the engineering gurus.  IBM with Watson is looking to take this a step further by bypassing the data techies and letting analysts explore data without as much consideration for the engineering/plumbing involved.  This is a similar direction that services such as AWS and Azure Machine Learning are heading, in the cloud.

In 2016 the biggest challenge for Big Data is turning down the focus on the technical how, and turning up the focus on the business driven why.  Engaging and educating those who understand a given business in the capabilities of data science, motivating them to lead these initiatives in their organisations.

About Tony Bain

Tony Bain

I am the co-founder of the RockSolid SQL business, the primary developer of the technology, and have built the business to include some of the largest and most well known customer logos.

My area of expertise is building solutions that deliver customer value via leveraging big data, machine learning, AI and software automation technologies. I have written numerous books, articles and posts on data driven business and have presented at conferences globally.

As a Director for RockSolid SQL I am responsible for:

  • Creation of the core RockSolid technology and technical innovation and development of the product
  • Data and analytics stratergy/implementation
  • Financial performance and growth
  • Customer satisfaction, retention and growth
  • Business development, sales and marketing and leadership on key opportunities
  • Service offerings, pricing and go to market strategies
  • Partnering relationships
  • Product strategy and direction for our core RockSolid technology
  • Technical leadership in key development initiatives
  • Team leadership and retention

The SQL/NoSQL war is over. The winner is… wait, was there a war?

SQL_NoSQL

We are approaching 7 years since the term “NoSQL” re-entered the popular tech vernacular, and 7 years since I wrote the post “Is the Relational Database Doomed?”. During this time, we have experienced a tidal wave of non-relational data management technologies. So, time for an update to my prior article.

At the start of the decade, when the NoSQL buzz was in its heyday, some were predicting the end of the dominance of the relational database platform (RDBMS) within a decade.  The reason for this seemed somewhat sound. That being, the relational database is based on what is now 40+ year technology and things are so much more advanced now than back then, so clearly this was a technology ripe for disruption.

So how has this disruption gone?  Well, all my metrics show there are more relational databases in existence today than at any point in history.  It may be hard for many people to comprehend the volume.  Often, mid-size enterprises operate hundreds of relational databases.  Many large enterprises have thousands to tens of thousands.  These represent the data stores of everything from ERP's, financial systems, content sharing apps, IT tools and so on.

So despite the noise surrounding NoSQL, in a head to head comparison of volume of use, NoSQL use seems so very small.  At a guess, I would predict that for every NoSQL database in existence there would be at least 1000 relational databases.  Probably more.  You would be forgiven for thinking NoSQL use was almost insignificant. 

So why has there been so little disruption?

  • The relational database has such a massive legacy. The IT world is full of people whose front of mind solution for a new data management requirement is a relational database.  To demonstrate this I looked on LinkedIn. My search showed that over 1,000,000 people list “SQL” as a skill in their profile.  In comparison only 16,000 listed NoSQL and 30,000 listed MongoDB.  That’s a massive skills gap.
  • The RDBMS is very general purpose.  There are very few day-to-day data management requirements that cannot be met without a run of the mill RDBMS.  So why would you not go with what you know, if what you know is suitable?
  • Relational databases are solving very complex problems in a balanced approach.  There is 40+ years of learnings on how to balance consistency, concurrency, scale and performance.  Many NoSQL initiatives focus on improving some objectives (such as scale or performance) at the expense of others (such as consistency or redundancy) solving their own problems but also lacking a general purpose appeal.
  • Relational database vendors have also kept innovating.  With some RDBMS vendors you can now combine SQL and XPATH, support JSON natively and support other non-structured data types.  Also, many RDBMS platforms now support in-memory databases and others are quickly adding this support.

So with the continued dominance of the relational database, what future is there for the NoSQL alternatives?  Well that is clear, the same opportunity as they have been filling over the last 7 years.  Edge cases.  Sure, enterprises have many routine data in/data out applications and these belong on the RDBMS but modern enterprises are trying to do more with data than ever before and leverage data in new ways for a competitive advantage.  Maybe they need data to be captured at a massive scale, much greater than what is possible with a traditional RDBMS.  Maybe they are looking to deep mine data to identify complex relationships between entities, or make predictions about how scenarios will transpire.  Maybe they are trying to learn from large and diverse data sets and discover key new ways to improve productivity.  This new world of requirements is where new world platforms have the opportunity to shine, focus on improving a specific set of key objectives, potentially at the expense of others, and then find the market that needs those objectives.

To summarise, I cannot see a world in the near future where any non-RDBMS gains any dominance in supporting the data management needs of most applications.  The vendors people use may change, and the location of those databases may change (on premise to cloud) but they will be relational. However, what are the edge cases of today will of course become more mainstream.  When this happens it will not be at the expense of the RDBMS, but instead they will be in addition to it.  The playing field is getting bigger.   Organisations desire to do more with data, via big data or data science initiatives that is fuelling a market ripe for vendors with clever, yet tightly focused, data management solutions.

So as it turns out, relational (SQL) and non-relational (NoSQL) technologies were not at war at all.  They are in fact allies, working together to deliver organisations both general purpose and special purpose data management solutions.