ADVICE: Read more, write less

Super interesting piece on the Savage Minds anthropology blog the other day titled Read More, Write Less, an essay by Ruth Behar about her conversations with the Cuban author and poet Dulce Maria Loynaz.

I must confess to having no familiarity with Ms. Loynaz, but in the piece she offers some really excellent advice for writers, bloggers, really communicators of any kind.

From the Savage Minds piece:

Inspired by her meditative Poemas sin nombre (Poems With No Name), I had written a few poems of my own, and Dulce María had the largeness of heart to ask me to read them aloud to her in the grand salon of her dilapidated mansion. She nodded kindly after each poem and when I finished I thought to ask her, “What advice would you give a writer?”

I will always remember her answer. It came without a moment’s hesitation and could not have been more succinct: Lee más, escribe menos, “Read more, write less.”

That might seem like old-fashioned advice in our world today, where so many of us aspire to write more. But having pondered Dulce María’s words, I think I now understand the significance of what she was saying.

It comes down to this: you can only write as well as what you read.

Awesome advice, and so good that I don't really need to add anything more to it. I try and read as much as I can in order to have new ideas, fresh perspectives, and just interesting things to share. But there is so much more out there.  I know I probably should read more, and different things instead of trying to push out posts all the time.

Read more, write less. Great advice. 

Have a great Thursday.

Learning by watching, something else at which the robots are superior

This story, Robots can now learn to cook just like you do: by watching YouTube videos, made the rounds over the past weekend. The basics of the story are these: researchers at the University of Maryland and an Australian research center have managed to create a system by which robots can 'learn' to cook, (how to recognize cooking tools, how to grasp and manipulate objects, how to process unfamiliar inputs into cohesive sets of instructiokns, etc), with the raw learning material consisting of a set of 88 YouTube videos of cooking demonstrations.

The entire paper, Robot Learning Manipulation Action Plans by 'Watching' Unconstrained Videos from the World Wide Web is here, but I will grab the most interesting and telling bit from the abstract, and then shoot a few comments after the excerpt.

From the paper:

In order to advance action generation and creation in robots beyond simple learned schemas we need computational tools that allow us to automatically interpret and represent human actions. This paper presents a system that learns manipulation action plans by processing unconstrained videos from the World Wide Web. Its goal is to robustly generate the sequence of atomic actions of seen longer actions in video in order to acquire knowledge for robots.

Experiments conducted on a publicly available unconstrained video dataset show that the system is able to learn manipulation actions by “watching” unconstrained videos with high accuracy.

There is a lot to unpack even in that short snippet from the research, but the implications of this research suggests a future state of even more powerful automation technologies - the kinds of technologies that can learn simply by watching. And unlike us puny humans, they won't get tired of watching the same stupid 'life hack' kinds of YouTube videos 73,000 times before getting frustrated that we can't 'get it' and then just giving up.

Some time back I posted about robot technology replacing or at least augmenting human staff in retail big box stores. In that post I posited that the real advantage, or at least one of the most important (and I think really overlooked for the most part), advantages that robots and technology have over human labor are the robot's incredible ability to learn, store, and share information with other robots.

When the robot solves, or learns how to solve maybe just by watching a human colleague, a customer's problem, it can instantly share that knowledge with every other robot, who will all then have learned to solve that problem. Information, learned knowledge then becomes an asset for all. Immediately.

Think about the power of that ability the next time you have to roll out some kind of training program to your entire workforce. How many times do you have to explain the same thing to another person? How long does it take everyone to 'get it?'

How many never do?

Twinkies or technology skills: Shelf-life is shorter than you think

All of us, or at least most of us that care about the future of some of America's most beloved snack foods, have been following the sad story of Hostess - the venerable maker of all things wonderful, (Twinkies, Ding Dongs, and my favorite, those orange nuclear-looking cupcakes), that after failing to come to terms with striking workers, appears to be in its death spiral.  If indeed Hostess has to liquidate, it seems likely that some of their most popular and iconic brands like the Twinkie might survive, with a number of potential rival baking companies seemingly eager to purchase (at a fire-sale price), the brand name and recipe for the cake.It might still be good

While a plausible scenario, the Twinkie's future is still uncertain, and in the week or so run-up from the initial announcement of Hostess' intention to pursue liquidation, and the last, failed attempt at a labor settlement, consumers across the country essentially bought out all remaining supplies of Twinkies and other favorite Hostess snacks. The idea being to stock up while you still could, and if you acted quickly and scored a few extra Twinkies boxes, that combination of your stockpile with the Twinkie's legendary decades-long shelf-life, you'd be set to get your Twinkie fix for a really, really long time.

But it turns out, despite the urban legend that the preservative and artificial ingredient-heavy Twinkie being able to last forever, (or near enough), the true shelf-life of a Twinkie is no more than about 25 days, and typically were pulled from store shelves after about 10 days. So bottom-line, Twinkies last for less time than you think, (please let's hope they come back, I finished my stash two days ago).

I thought about the little Twinkie paradox while reading this piece, 'What's the Shelf-life of a Techie? Just 15 years, from the Times of India site.  In the piece, high-ranking technology leaders from the India operations of several well-known global tech firms,  (SAP, Microsoft, Texas Instruments), paint a pretty stark and probably realistic picture of the increasingly rapid deterioration of technology skills and expertise, as the pace of newer, hotter, and more in-demand technologies come to market. There are about half a dozen choice quotes in the piece, but this is the one that really stood out the most:

Mukund Mohan, CEO of Microsoft's startup accelerator programme in India, says the shelf life of certain kinds of developers has shrunk to less than a year. "My daughter developed an app for iPhone 4. Today, she is redeveloping the app to make it smarter for iPhone 5.

Five years ago, developers were talking Symbian (the Nokia operating system). Today, it's not very relevant. You have to look at Android or iOS or may be even Windows 8 to stay relevant."

A pretty telling quote and a bit frightening as well. The iPhone4 is maybe a year old, the iPhone5 less than that, and about to be rendered, 'out of date' in maybe 6 more months. It used to be that hardware, software, and the technical skills needed to make it all work advanced more evenly, regularly, and more importantly, the big firms that make these technologies, and their customers that use them, had much longer time horizons in mind when developing and deploying technology.

In the (recent) past a large, enterprise deployment of an ERP or an accounting system at a big firm could reasonably be expected to be in place more or less unchanged, for a decade, maybe even longer. Lots of IT pros and managers have made long and successful careers essentially by developing a deep understanding of a single technology.

While that was commonplace, and if you chose the 'right' technology, a pretty shrewd approach to career management, if these IT executives are accurate in their assessments, specializing in one technology at the expense of, exposure to, and continual learning about the 'next' technology that will be in demand is the only way to at least have a chance of remaining relevant, (and employable), past a time horizon, like the shelf-life of the Twinkie, is a lot shorter than you think.

Have a great Tuesday!