Conning America

Someone really should write a book on Trump and how he is conning America.

Conning, in both senses of the word "con."

To con a ship means to be at the helm and to steer it. Which is befitting Trump since as the president, he is the man in charge of steering America.

There is another meaning for "con" for which Trump appears to be doing, with his campaign promises and what he has delivered to date.

So "Conning America" is going to make a great title for a book about Trump's presidency. You can even add in "The truth about the Trump presidency" as a sub title for the book.

All we need now is someone to pick up on this, do the research, write the book, and get it published.

Maybe Stephen Colbert or Trevor Noah?

#ConningAmerica #StephenColbert #TrevorNoah

Horrified by tweet

The morning started with this horrifying set of tweets.

 
I don't know what medical costs he is talking about, but I don't think being transgender affects how decisive a person is. And opening serving as transgender takes a lot more courage than what most people have.

If you want decisive and overwhelming victory, you should be looking at how to attract the best people and how to keep them, rather than worry about what's between their legs. Kristin Beck is transgender, but that did not stop her from being one of the best military personnel, making it to serve on the Navy SEALs. There are many more serving now who are just as effective as any other trained military personnel, and the least the top guy can do is not distract them from their duties.

If it is about medical costs, well, there are a lot of other conditions that require medical attention. Does that mean all these people are banned from serving too? According to some sources, transgender people serving in the US military may cost the military up to 8.4 million USD annually. Compare this with the 41 million USD spent annually on Viagra.

So when will people wake up and do the right thing to save the US. I hope it is soon, before it is too late and there is nothing left to save.

A new industry of fact checkers

Ever since Trump has entered the news running for president, there has been a whole new industry of fact checkers. No doubt there had been fact checkers in the past. It is a part and parcel of democracy for people to check if what someone said is true, and to expose any untruths. That is what free speech is about too. And therefore, people do check what they say beforehand so that they don't risk having someone come out and saying "that's not true."

But ever since Trump came into the picture, with so many twists and turns and what have you, it seems the usual number of fact checkers just can't keep up. Mainstream media has to get teams of fact checkers to run through everything just to make sure what they report are factually correct. And people are taking to social media to bring out any untruths too, bypassing the mainstream media (which, no matter how hard they work, is still limited by the size of their budgets). We now have YouTube channels dedicated (almost) to exposing such untruths.

If anything, he has provided employment for some people: fact checkers. But fact checking is not productive work; it doesn't produce anything new. It will be better for society if such resources are channeled into other areas. But I guess as long as people are not as truthful as they should be, or as thorough in their own checks before speaking out, fact checkers will continue to be necessary, and this industry's size will depend on how prominent (or how large the scale) of such behaviour.

My simple answer to the grandest staircase problem

For those who have heard of the_grandest_staircase_of_them_all problem, this is my simple solution. However, it is not my final solution. My final solution is an optimized version of this. The bigger the value of n, the longer this version will take, since it is recursive and the recursions become deeper and wider when n gets bigger.

def answer(n):
    def times(x, y):
        count = 1
        for k in range(x + 1, y // 2 + 1):
            if k < y - k:
                count = count + times(k, y - k)
            else:
                return count
        return count

    counter = 0
    for i in range (1, n):
        if i < n-i:
            counter = counter + times(i, n-i)
        else:
            return counter
    return counter

Hope is hope... not!

When my boss tells me, "I hope to see the report tomorrow," I take it to mean that I better make sure the report is on his desk tomorrow or I will find myself without a desk soon.

When the mafia boss tells his underlings, "I hope that pesky reporter has an accident," his underlings are not going to walk away and drink beer. They are going to go and fix something up to make it look like an accident.

When wife tells husband, "I hope to have ___ for my birthday," boy, he better get ___ for her for her birthday.

Hope is hope only when there is no power relationship between two persons.

Japan offers to work with China on Silk Road initiative

Abe offers conditional cooperation with China's Silk Road initiative

China is probably laughing at this.

The Silk Road Economic Belt and the 21st-century Maritime Silk Road, also called The Belt and The Road in short, uses the name "Silk Road" because of its historic significance. The Silk Road linked China to India and Europe in the past. This led to the flourishing of trade and culture along the route. And historically, China has held influence (big or small, depending on China's domestic situation at each point in history) over states along the route. China had always sought to maintain security along the route, for the benefit of the trade and its positive effects on China's economy.

So naming this new initiative "Silk Road" is a blatant way of saying that it is for China's benefit, and it is going to be a way for China to influence nations along the route (actually, two routes; one land, one maritime). And in the past, states not directly on the route could still benefit indirectly if they bowed down to China. Japan was no exception; when relations were fine with China, Japan was able to obtain goods from Europe and India that came through the Silk Road to China.

Make no mistake: the new initiative is for China's benefit. Not solely; if you happen to be along the route, you will benefit too. But if you are not on the route, and you want in, you will only be included if it is to China's benefit. That is to say, you can only get in if you bring something to the table that China wants (like in the old days, when vassal states brought tribute to the Chinese emperor). So unless Japan can bring something that China can't resist (maybe some of the rare metals that they just found off Chiba)... well, China will just look at Japan and laugh.

PowerPoint makes us lazy

Or at least, that's what I think.

Because in PowerPoint, we usually write in point-form, and not full sentences. And we usually only write in the key points, without filling in the details on how we arrive at them, or how they link or lead to each other. All those details are usually verbally explained during the presentation (if there is one) or simply left for the reader to figure out (if there isn't an actual presentation and only the slides are sent out.)

This is not a problem if the slides are created based on a full document, properly written with all the details on how things were arrived at, and how they link to each other, etc. The slides are a summary of the key items, and readers can (and should) always refer to the full document to get a comprehensive idea of the issue being talked about.

But it is a problem if the writer starts right from PowerPoint, generating just the slides. While the key points are there, the writer might not have fully developed the idea, might miss out important (or non-critical but important) factors or aspects that would have been explored if he had started by writing the issue out as a document (be it essay, proposal, research paper, or even just as an email). Because grammar forces us to consider the relationship between objects and subjects, think about time (past tense, present tense, etc.), and how to link things together.

Long story short: use slides for what they were meant for, as a summary of a document. Don't be lazy, write the document, then develop the slides to help others better understand the document.

U.S. leaving the Paris Climate Agreement

And there you have it.

Trump has announced that the U.S. will be leaving the Paris Climate Agreement. All because it will cause the U.S. to lose jobs.

If jobs are lost because companies are not environmentally-friendly, I think the onus should be on the companies to comply with climate rules even as they take on production. Change their methods of production, look into what they are using for raw material, you know, make an effort. It may be a bit more costly, but in everything we do, there is a cost. That's the basics of economics. Companies either bear the monetary cost of environmentally-friendly production, or society will have to bear the social cost of pollution.

Personally, I think these rules are good. They force companies to think about the environment, so that they bear the monetary cost of production, rather than passing it on as social cost to society as a whole. Companies will find ways to be profitable no matter what the rules--necessity is the mother of invention. Instead of allowing companies to grow lazy and still reap profits at the expense of the environment, we should keep these rules to protect our environment, since companies will find ways to profit anyway.

Oh well, I am not the leader of any country, so I guess what I think or say doesn't really matter. Time will tell if decisions are sound. I just hope we still have an Earth by that time.

Generalists and specialists

As artificial intelligence develops to become even better at learning what humans can do, I think it is even more important for us humans to think about what we should do. What are the things that humans are good at, and what are the things that will eventually become things left to machines?

First, the success of AI in learning single tasks, and the greater difficulties with integrating things from a wider range of fields mean that humans will probably continue to have to be the generalists that binds together different fields of specializations to achieve an overall bigger objective. So we probably will continue to need managers with people management skills able to bring together a team of people with different talents to work on a project.

But how about specialists, people who are good at doing certain things? Well, the recent success of deep learning has shown that machines can now be trained to do certain things much much better than humans. That's because people get better at these things by doing them, and machines are able to devote a lot more time to doing this things repetitively to become better at them than humans. So machines are likely to replace cucumber sorters, since they can learn to do this offline (by training on pictures of cucumbers, which requires a lot less time than actually finding cucumbers to look at and sort), and once trained, they can be put to the actual sorting task for hours and hours without having to worry about labor laws about working hours. Compared to humans, they take less time to train, and can work for much longer hours.

However, we will probably still need specialists in which ability is not just gained by repetitive learning, but also requires some form of creativity. Although there is on-going work to teach machines creativity, based on deep learning techniques, these are still limited to training using a big dataset to try and generate something from it. It is evolutionary, not revolutionary. For that revolutionary break, we still need human creativity, and that is where humans will continue to find a niche for them to specialize in.

In short, as computer become better and better at mimicking humans, through advances in AI, humans will need to find a place for themselves. These could be as generalists, where AI is still unable to adequately mimic, or as specialists in fields requiring human creativity.

Two types of wargames

Recently, I have been reading a book about Japan's defeat in WWII. It led me to think about wargames in general.

To me, there are two types of wargames.

One is the confidence building type. The training type. Where the enemy behaves in much the way that current intelligence assesses he may behave. It is about executing a plan against a known enemy, to test the effectiveness of the plan, build confidence in the plan, and train people in how to carry out that plan.

It is the "winning" type of wargame. Where the good guys (the friendly side) is supposed to win. The enemy behaves like how we think he will behave (which may be close, or way off, depending on how good our intelligence and assessments are).

The other type of wargame is the "losing" type. This is when the enemy does what he can do, given the resources available to him. The difference is between can and may. "May" is about probability, about what he is likely to do. "Can" is about possibility, no matter how small, and is about what he is able to do. Here, the goal is to try to beat the good guys, to make the friendly side lose. This is to expose the weaknesses in existing plans, to expose areas which have not been considered in planning or were not answered. Here, consideration must be given beyond the military to bring in economic and politic aspects of war, since war is not a standalone condition and is part of a bigger grand strategy.

It is about identifying risks, military or otherwise, so that they can be dealt with.

But nobody likes to lose. Nobody likes to be shown his weaknesses. And it can be bad for morale too. So this type of wargame is rarely conducted. Yet it is just as important as the "winning" type. Only with both will you be able to be fully prepared for this dangerous feat known as war.

So I think what is important is that in a wargame, we must not be afraid of losing. Because losing helps to make the plan better. And it is better to lose in a wargame so that we can improve the plan, rather than lose in an actual war, by which time we won't have the luxury of time to improve the plan.

Facebook's grammar is...

Just when Facebook announced this new phase in their machine translation research...

A novel approach to neural machine translation

This came up in my Facebook feed.

"Rain is forecast."

Somehow, it doesn't seem right to me. I would think it should be "Rain is in the forecast" or "Forecast is rain"... I mean, we say "sky is blue" but not "blue is sky", right?

Then it got me thinking. Maybe Facebook got their forecast for Yokohama from a Japanese source (which only makes sense, right?). Which means it was in Japanese. Which means it was then translated automatically by Facebook's machine translation software into English to put this on my feed. Which means this shows the current state of Facebook's state-of-the-art machine translation for Japanese to English.

I guess I still have time to earn a living as a translator.

We need fats to give birth to the next generation

A thought that came to mind a while ago.

Humans and companies are quite alike in a certain way. Both need fats in order to be able to give birth to the next generation.

For humans, females need a certain level of body fat in order to have normal reproductive functions. When there is not enough fat, reproductive functions cease, which means they cannot give birth to the next generation.

Similarly, companies need fats (not the oily kind, but as in having some form of abundance or excess of something) so that they can devote part of their resources towards training and education to nurture their next generation of people. If companies are so caught up in the day-to-day work that they have no time or money to put into nurturing their people, they will not be able to sustain their operations.

While this is a simple analogy, I think it shows how important having some form of abundance is. This even shows itself in artificial societies. In the Sugarscape model, agents need to gather a certain amount of resources before they can "reproduce". This is the "excess/abundance" that we see in the real world, whether biological or financial. A good reminder for myself in any future research into artificial societies.

Star Wars Day - 40th anniversary

Today is Star Wars Day!

This year marks the 40th anniversary of Star Wars. Wow! 40 years!

May the Force be with you!

#Maytheforcebewithyou #StarWarsDay

Beauty and the Beast (2017 film)

An incredible story. An amazing story. A story that has continued to capture my heart since the day I first saw it in 1991. The live remake of the 1991 animated film adds to the story by trying to clarify Belle's and the Beast's childhoods, although I felt that these parts weren't really necessary since it was already a great story to begin with.

What really captures me is the story, about not being afraid to live life as you are (even if you are the only girl in the village who can read) and to look beyond appearances ("for who could learn to love a beast?"). Both the animated film and this new remake combines a great story with great visuals and amazing music.

Yet another great work by Disney that will touch your hearts and bring tears to your eyes.

#BeautyandtheBeast

Some thoughts on artificial intelligence

Enabling Continual Learning in Neural Networks

This is an article about a month ago from DeepMind. It does seem that we are getting a bit closer to being able to use artificial neural networks (ANN) to replicate the learning that takes place in the human brain.

And I think that is where the problem lies.

We are trying to replicate the human brain, but is the human brain the best model for learning?

An example. My eight-year-old son and I both watch the same Star Wars movie together. Yet he can recall a lot more details (the scenes, the people, even the music) about the movie compared to what I can recall.

If what DeepMind is doing is a hint, it may mean that our brains are also trying to optimize the limited synapses (although the number of synapses in the brain is way beyond what any ANN has now) in the brain. Which means that once we have used our brains to learn a certain amount of things, any additional learning is going to mean readjusting the connections (weakening our memories in some areas) to learn those new things.

Going down this road, it suggests that making machines think and learn like humans may not be optimal, as the way our brains learn has a serious flaw--it needs to weaken our ability to recall what we have already learnt, in order for us to learn new things.

If only there is some other way to learn things without having to forget what we have learnt. Well, for our mortal brains, that may not be an option, but it is something that we can try in our attempts at giving intelligence to machines. And I think that is the Holy Grail of artificial intelligence--to arrive at a model for machine learning that surpasses the human brain.

#DeepMind #AI