Category Archives: future

Alexa and Her 10,000 Skills

First, the news. Alexa toolbox of skills is growing at an accelerating rate.  From Wired

While 10,000 may seem like an arbitrary milestone, it’s an instructive one, especially when you consider how fast it’s come. Last June, a full year after the ASK launched, Amazon announced that Alexa had reached 1,000 skills. By September, that number had tripled. In January, Alexa’s skills catalog swelled to 7,000. It took just over a month to tack on another three thousand.

Unfortunately, as Alexa becomes more skillful, it becomes more difficult to figure out what Alexa can and cannot do. In other words, Alexa is not getting much smarter.

Sadly, I think that we will be stuck in this paradigm for a while. The reason is that the skills that Alexa is getting generally link us to suppliers of pre-fab solutions. So, for example, Alexa can adjust the oven temperature for a GE device as needed. Alexa can order a pizza, as long as it is form Domino.

Using Dan Kahneman’s line of thinking, this enables us to do things faster. It does not enable us to think of better ways to do things. Can we get better ideas from AI powered bots? Of course we can. But to get that, we will need a paradigm shift. Instead of the AI  imprisoned by a given supplier, like GE or Domino, the AI needs to be free to serve our needs. So when we ask Alexa a question, Alexa should give us the BEST possible answers, not just the one that a given company wants us to hear.

And when will this happen? The first step is not technological. It is devising a new business model where we can make money by giving and getting best possible solutions to questions — solutions that open the door to further discussion, not just a one off transaction, like ordering a pizza.

Will this happen? I think so and when it does, it will shake up the world as we know it.

Leaving Donald Trump Behind

Would it be fair to say that Donald Trump is the last gasp of 20th century man? That might sound a bit odd, but if you think about it, Trump does embody certain characteristics that are sooooo last century.

The most important one is certainty. Trump is certain that he is right. It matters not what others say or think. Trump believes that he is smarter and has the right answers. Indeed, he distorts reality when he needs to in order to justify his opinions.

So how is this a 20th century phenomenon? Consider that in the 18th century, the enlightenment opened the door to speculation that was not God centered. Western man started to think in terms of science rather than religion. And this led in the 19th century to certain “scientific views” about history. I refer to Hegel and Marx in particular. In the 20th century, these so called scientific views hardened into ideologies – communism, fascism, capitalism. Each ideology was certain that its view of history and the world was correct. Ironically, the tool that opened our minds — scientific method — then closed them. Amidst mind boggling scientific advances, we clung to the certainty of our views on the most basic issues.

And of all the unpleasant character traits that Donald Trump exhibits, his closed mindedness is perhaps the most obnoxious.

That tells us something. The 20th century is over. We are now embarked on a new century that will develop new ways of seeing the world. If we want to visualize what we would like to leave behind, just think of Donald Trump.

The 21st Century Mantra

My dream is that by the end of this century (2100) humanity will have fundamentally altered the way that we live in a very particular way.

To bring out what this change might look like, let’s ask a question about he past. How much were we dedicated to learning? The answer — in very general terms —  is that learning took second place to surviving. Surviving from day to day was a challenge for most of our history, which meant that most of humanity focused less on learning, especially learning for learning sake. The result? A less than optimal learning path. We could have progressed further than we have in understanding the reality that surrounds us.

There are very entertaining stories that play on this theme, for example Mark Twain’s, A Connecticut Yankee in King Arthur’s Court.

Image result for Connecticut Yankee King Arthur

Things have changed, at least to a certain degree. After the 20th century, many more of us (1) enjoy the luxury of deciding what learning paths we will pursue before entering the work force, (2) enjoy access to a stream of information about what is going on around us, and that from time to time helps us learn, and (3) rely on learning to add value to society.

All of that is progress. But we can do better – much better. We can do better if we understand how learning adds value. What is value? We tend to see value in terms of exchange value —  something is worth what someone else will pay or exchange for it. In other words, value is a a social rather than individual measure. And so when we ask how does learning add value, we are asking how does learning add value in terms of exchanges.

Here is the key idea – our institutions are pretty good at delivering value based on present knowledge. They are not organized to optimize introduction of new knowledge into products and services. Why? Firms make money by selling what they make NOW, not what they MIGHT MAKE in the future.  Indeed, it was expensive to invest in the technology that enables them to produce what they sell now. They have a strong incentive to continue using it as long as possible without investing in new technology. Markets can force them, and that is better than life without markets. But markets acting alone cannot push firms to optimal levels of innovation. If you doubt this, consider that government supported research has financed just about every major innovation that we now enjoy. Greg Satell writes

Take a look at any significant invention today and much of the core technology began with a government grant.

We can do better, if we understand a very basic idea. Learning adds value in terms of exchanges through 3 steps

  • discovery of new knowledge
  • adapting the knowledge to address a shared problem
  • developing products and services based on the adaption

We might add two additional steps. The first one precedes acquisition of new knowledge – it is developing platforms that build the capacity to discover new knowledge. The second one is at the end – developing platforms that track and share the learning paths that make progress possible.

In other words, if we want to maximize the value of learning in society, we need to invest in each of the above five steps.  Each of these occurs in different sorts of groups. But the groups need institutional connections so that they can exchange ideas and build a common vocabulary.

Societies that can make those interconnections work better will thrive in the 21st century. By the end of the century, let’s hope that we all are organized around this pathway to adding value through learning.

Fight for Your Future!

Yesterday, I posted on Al Wenger’s idea for re-shaping politics. It is not about left versus right. It is about past and present versus future. right on Al!

The key point is worth repeating. The future that we want, where innovations address and solve the pressing crises that face mankind and where our kids live far better lives than we could ever imagine will not happen automatically. There are too many interests at work that hold us back, either pulling us back to the past or seeking to freeze us in the present.

It is time to think about fighting for your future. If you don’t demand it, you won’t get it!

Re-Framing the Debate

For some time now, I have been struggling to come to terms with the political mood shifts that characterize our politics. It is not just about Donald Trump. In fact, my concern started  around the time that Vladimir Putin came to power in Russia.

My problem was in finding the right way to characterize the underlying conflicts that was at work. Al Wenger has done this very well for me. Al writes

Many of the people in power or currently grabbing for it are trying to maintain the recent past or even go back further. This is the fight of our lifetime. The Past against the Future. It is NOT: left against right, rich against poor, black against white, Islam against Christianity, men against women.

Right on!  We have the tools to build an amazing future for all of us. But we may do something very different if we remain trapped in the vocabulary of the past.

Thanks Al! Onward!

Preparing for the 2018 Recession?

I know. 2017 is just getting started and we know very little about what it will bring. How can we know what 2018 will bring?

We might offer a few guesses by looking at longer run data. If we can identify longer running trends that have started earlier, we might extrapolate them into the next several years. Yes, this is uncertain. It is also important because at least some of these extrapolations trigger alarm bells.

Consider this one

The old, industrial era rules for the dying age of energy and technological super-abundance must be re-written for a new era beyond fossil fuels, beyond endless growth at any environmental cost, beyond debt-driven finance.

This scenario posits that the demand for fossil fuels will not fall even as supplies tighten. As a result, the global economy will experience a price shock for energy that may trigger a recession.

Will this happen? Who knows? But the fact that it might happen should trigger concern about where we are now and what we should be doing to avoid the worst.

Are physicists taking over Silicon Valley?

If you are like me, you want to know how to succeed in the future. What things can I learn how to do now that will pay off over time?

If this question interests you, consider this from Wired

… this is a particularly ripe moment for physicists in computer tech, thanks to the rise of machine learning, where machines learn tasks by analyzing vast amounts of data. This new wave of data science and AI is something that suits physicists right down to their socks.

This is so because physicists are trained to deal with enormous amounts of data, as well as complex math. That might now be your bag, but there is one more aspect that may suit you more.

Physicists are trained to deal with uncertainty. And that is a key area to understand deep learning.