Tagged: Data science
Mini review: “The Rise of the Robots” by Martin Ford (audiobook version)
The impact that exponential technological advance will have on the future of the economy really came into focus for me five years ago when I read “Race Against The Machine” and Martin Ford’s first book “The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future” in quick succession. The two were both were very cheap, unprofessional looking e-books and the writing could be improved. They were pedalling ideas that were outside of mainstream thinking. People where just coming to terms with the bombshell that was the Great Recession.
Fast-forward to today and the ideas have come to the attention of the general public. Back then we had a chess champion beaten by brute-force computing and IBM’s Watson beating humans at a quiz show. Now we have self driving cars, a Go champion beaten by a computer program using machine learning and dire warnings from important figures in science and technology warning of the risks of artificial intelligence to the future of the human race.
Also, in the meantime both books have had an upgrade and big improvements in quality. “Race Against The Machine” became “The Second Machine Age” and “The Lights in the Tunnel” became “The Rise of the Robots”.
Although the message is largely the same as the earlier book, the author’s writing is much improved, the arguments are clear and supported with data, there is little repetition in subject matter and the book is kept short (350 pages / 10 hours).
This book won the prize for Best Business Book of the Year 2015 and it is easy to see why. However, I do wonder – as this book has come to the attention of the business community – whether a large part of the book’s audience will actually take heed of what is being said especially in regard to the issues of income inequality and productivity if large scale automation is undertaken. They certainly cannot now say that they have not been warned. Will business leaders believe that the message applies to others but not to themselves?
A note on the narration as I listened to the audiobook version: sometimes these business books end up being read by a default-American-voice which can sound very bland and unenthusiastic. In the worst cases, this can actively detract from the content of the book (The Second Machine Age is a victim in this regard). When I heard the voice of the narrator for this book I was initially concerned. In the end Jeff Cummins does a good job, adding intonation and a relaxed tone which matches the style of the text.
In conclusion, this book should be read by everyone. I would go as far as saying that the book is an addictive read (or listen). If you have not thought about the subject before, it will really give you something to think about. You may be persuaded at least that the author has a point or you may be convinced he is right. He sounds right to me.
Next steps towards Data Science
After my last Open University course (Analysing Data) ended in June, I gave myself the summer off. I read a number of books and enjoyed the time that was previously taken by study with my family. I decided that the time demands of another OU course (along with work commitments) would be too much at the moment.
But that does not mean that I can’t carry on towards my goal of moving into the field of data science – I just need to approach it at my own pace.
There are a few new books that are starting to get to grips with what is a new and largely undefined subject, especially “Big Data“, “Doing Data Science” and “Learning R” . I feel that people can now become a lot more informed about what is involved. There is now more flesh on the bones. From here on in I’ll know what is actually involved and what is needed to get there. The more I do and the more I find out, the more challenging it seems. Daunting even. But I firmly believe this is a foundation for the future and I fully intend to take part, even if there is a long way to go.
So, what’s next? These are things that can be done concurrently. I intend on doing more practical work.
• Continue to learn Python – I’m getting to grips with the basics
• Read “Doing Data Science” – I’ve started, and I think it will be a real education on what is really involved
• Start “Learning R” once “Doing Data Science” is finished, although Python will be the focus for the near future.
That should keep me busy for a while…
Review: Big Data – A Revolution That Will Transform How We Live, Work and Think
The bottom line first: if you are interested in a broad and non technical introduction to the subject of “Big Data” then you should read this book. It is short and highlights a number of points (some that aren’t necessarily clear from reading elsewhere.)
Point 1:
Importantly in the first chapter it says that to be practising “big data” projects you do not have to be dealing with millions of data points. There may be a lot less but the issue is that you should be working will all the data that is available to you rather than just a sample. With all the data, it is possible to analyze it in different ways. With just a sample you will likely be limited to what you can discover after the sample has been taken. The authors discuss the very first article I read about this subject, Wired’s The End of Theory. It’s very interesting to read how the article is now regarded.
Point 2:
People may have to get used to the data revealing what is happening without actually revealing why it is happening. In some areas we will have to let go somewhat of the (natural) desire to understand the reasons behind the results.
Point 3:
The authors deal with the subject of data getting “messier” (becoming more imprecise) as as you increase the amount you are collecting:
However in many new situations that are cropping up today allowing for imprecision – for messiness – may be a positive feature not a shortcoming. It is a tradeoff. In return for relaxing the standards of allowable errors, one can get a hold of much more data. It isn’t just that “more trumps some” but that, in fact, sometimes “more trumps better”.
Because this data set consists of more data points, it offers far greater value that likely offsets its messiness.
Big Data transforms figures into something more probabilistic than precise.
So more trumps less. And sometimes more trumps smarter.
“Simple models and a lot of data trump more elaborate models based on less data.” (quote from Peter Norvig, Google)
… treating data as something imperfect and imprecise lets us make superior forecasts and thus understand out world better
Point 4:
The chapter on “Datafication” of just about everything is a good balance of history and the insights that can be gleamed from today’s social media giants. Location is particularly important:
The point is that these indirect uses of location data have nothing to do with the routine of mobile communications, the purpose for which the information was initially generated. Rather, once location is datafied new uses crop up and new value can be created.
Datafication is only just starting, but now it is under way it will continue, with many benefits:
Once the world has been datafied, the potential uses of the information are basically limited only by one’s ingenuity.
Seeing the world as information, as oceans of data that can be explored at ever greater breadth and depth offers us a perspective on reality that we did not have before.
Point 5:
Another important point is that humans will have to get used to the fact that their opinion is not always the best:
… the biggest impact of big data will be that data-driven decisions are poised to augment or overrule human judgement.
This is likely to mean a change in the requirements needed to do a specific job. The importance of experience will diminish as insight from data can dwarf the experience of one person.
Mathematics and statistics, perhaps with a sprinkle of programming and network science, will be as foundational to the modern workplace as numeracy was a century ago and literacy before that.
… the winners will be found among large and small firms, squeezing out the mass in the middle.
Big data squeezes the middle of an industry, pushing firms to be very large, or small and quick, or dead.
Point 6:
Re-use of data is looked at – old data can be combined with new in different ways to discover or exploit new opportunities. So what is the value of data? A company may have relatively few assets but a massive company valuation – therefore is the difference between the two the value of the data the company controls? That could mean billions of pounds / dollars / etc.
And finally…
A number of times there were names of sites or companies that led me to put the book down, check out a website or install an app. The chapter called “Implications” is particularly good for that, but it does slow down the reading somewhat. Even when a book is this recent some of the examples are now out-of-date (for example, Decide.com shutting its doors as its staff join ebay). This is a fast moving field.
There is a lot more to this book, impressive given that it is only 200 pages long. I’m glad I read this book – it puts so much into focus.
Spare Cycles 