Showing posts with label Google. Show all posts
Showing posts with label Google. Show all posts

Monday 16 May 2022

Nine Cool Ways to ‘Rethink Thinking’

Detail from the cover illustration for the book Rethinking Thinking


By Martin Cohen

I’ve been thinking lot about thinking a lot! Here’s some of my thoughts…

Rule Number One in thinking, via the classic text The Art of War, is don’t do things the clever way, nor even the smart way: do them the easy way. Because it doesn’t matter what you’re wondering about, or researching or doing - someone else has probably solved the problem for you already. Flip to the back of the book, find the answer. In fact, the great thing about the ancient Chinese book, The Art of War, is that it is not a brainy book at all. It is really just unvarnished advice expressed in what was then the plainest language. We need more of that. Thinking Skill Two is to avoid ‘black and white’ thinking, binary distinctions, ‘yes/no’ language and questions, and instead interact with people in a non-linear, less ‘directive’ manner. Take the tip from ‘design thinking’ that approaches rooted in notions of questions and answers are themselves limiting insight, because questions and answers are like a series of straight lines, when what is needed is shading, colours and, well, ‘pictures’. This is why it is sometimes better to go for narratives – which are conceptually more like shapes.

Tip Number Three, which, yes, is connected to the previous tips, and that’s a good thing too, is to look for the pattern in the data. (That’s what Google does, of course, and it hasn’t done them any harm.) More generally, as I say in chapter three, these days, scientists and psychologists say that humans beings are, fundamentally, pattern seekers. Every aspect of the world arrives via the senses as an undifferentiated mass of data, yet we are usually unaware of this as it is presented to our minds in organized form following an automated and wholly unconscious process of pattern solving. However, there’s a caution that has to come with advice to pattern match, because as we become attentive to some characteristics, we start to latch onto confirmatory evidence, and may neglect - maybe indeed suppress - other information that doesn’t fit our preconceptions. Put another way, patterns are powerful tools for making sense of the world (or other people), but they aren’t actually the world – which may be more subtle and complex than we imagine.

Thinking Skill Number Four, coming in quite a different direction, is ‘pump up the intuition’. Instead of thinking about ‘what is’, think about what might be. The ability to create imaginary worlds in our heads is perhaps our most extraordinary mental tool –yet so often neglected in the pursuit of mere observations and measurements. The ability, that is, as I explore in chapter four, to take any set of facts and play with them, to consider alternatives and hypotheticals. Being ‘playful’ is what the approach is all about, and something that probably the techniques’s greatest exponent – Einstein – stressed too.

My Tip Number Five actually looks more of a caution than a counsel: put all your cozy assumptions to one side and instead be ready to rigorously test and challenge every aspect of them. This is part of the solution to the age-old problem of 'thinking you know’, when, in fact, you don’t, which has been the task of philosophy ever since Socrates walked the streets of Ancient Athens challenging the arrogant young men to debate. 'Thinking you know everything' about an issue is invariably intermingled with the tendency to only see what you expect to see, likely because you have got caught up in a particular story. Exploring some of the stories that make up the amazing tale of the moon shows how this, ‘engineering’ approach, an approach we wrongly dismiss as rather dull, actually has enormous creative power. Socrates understood that: demolishing old certainties is not an end in itself, but a precursor to allowing new ideas.

Problem Solving Strategy Number Six, my favourite tip, is to doodle! Doodles, it turns out, are another kind of ‘intuition pumps’ (like thought experiments) and that gives you the clue as to what is valuable about doodling. It is emphatically NOT about the visuals – yes some people can draw beautiful doodles, many people to curious ones, and some of us do downright awful ones – but the value of doodling lies in the freedom it gives your subconscious mind - with apologies to all those specialists in psychology who will insist on some technical definition of the term. Ideas, as the Japanese designer Oki Sato says, start small - and can easily wither at the first critical assessment.

Doodling can be metaphorically compared to watering a thousand seeds scattered into a plant tray. You don't know at the outset which of the tiny seeds will be important - and you don't care. Instead, you give them all a chance to grow a little. Doodling, we might say, is similarly non-judgmental, and, at the same time, actively in praise of the importance of small thoughts. Of course, we do this all the time here at Philosophical Investigations!

Strategy Number Seven is about a very different kind of thinking. It's all about finding the explanatory key to make sense of complexity - the kid of complexity that is all around us at every level. Here, the thinking tools you need are an eye for detail along with an ability to see broader relationships. Put that way, it seems to require rather exceptional abilities But the thinking tool here is to stop trying to work out, and maybe even control a complex system from above. That's the conventional approach – where you gather all the information and then organise it.

No, the tip here is let the complexity organise itself - and you just concentrate on spotting the tell-tale signs of order. Watch for the 'emergent properties' that arise as a system organises itself and devote yourself to preserving the conditions in which the best solutions evolve. Put another way, if the cat wants food, it will keep pestering you. Be aware of the patterns in the data!

Tip Number Eight, is the computer programmer's cautionary motto: ‘Garbage In, Garbage Out’. To illustrate this idea in the book, I looked at the science governing responses to the corona virus, which was indeed heavily driven not by real-world data, but by computer models. But forget that distinction for a moment, the point is really a very general one. In everything we do, be aware of the quality of the information you are acting on. It could be compared to decorating a room: before you apply the colorful paints, before you lay the expensive carpet, have you prepared the surfaces? Mended the floorboards? Because shortcuts in preparation carry disastrous costs down the line. It's the same thing with arguments and reasoning. That dodgy assumption you made in haste, or perhaps because checking seemed time-consuming (or just boring) to do – can easily undermine your entire strategy, causing it to fall apart like at the first brush with reality. Just as an elegant carpet is no use if the floorboard is creaking.

Thinking Strategy Nine, my last one here and the one ever so slightly paradoxically closing my book, is to use ‘emergent thinking’, by which I mean adopting strategies in life that involve exploring all the possibilities and then generating new ideas, concepts and solutions from out of combinations of those ideas which likely could not be found in them individually. Yes, the strategy sounds terribly simple, but that first stage - of exploring possibilities - can be terribly time consuming and literally, impractical. So there are a host of micro-strategies needed too, like learning how to find information quickly, and above all how to select just the useful stuff. This is why I call this strategy ‘thinking like a search engine’ in the book, but, as I explain there, search engines actually rely to a large extent on human judgements: about information relevance and quality.

You can't get away from it: sorting the garbage requires a little bit of skill along with the apparently trivial, routine procedures. Thinking requires a brain - whatever some researchers on mushrooms may tell you…



*Martin Cohen’s book Rethinking Thinking: Problem Solving from Sun Tzu to Google was published by Imprint on April 4th.

Monday 9 May 2022

Peering into the World's Biggest Search Engine


 If you type “cat” into Google, sone of the top results are for Caterpillar machinery


By Martin Cohen and Keith Tidman


How does Google work? The biggest online search engine has long become ubiquitous in everyday personal and professional life, accounting for an astounding 70 percent of searches globally. It’s a trillion-plus-dollar company with the power to influence, even disrupt, other industries. And yet exactly how it works, beyond broad strokes, remains somewhat shrouded.

So, let’s pull back the curtain a little, if we can, to try observing the cogs whirring behind that friendly webpage interface. At one level, Google’s approach is every bit as simple as imagined. An obvious instance being that a lot of factual queries often simply direct you to Wikipedia on the upper portion of the first displayed page.

Of course, every second, Google performs extraordinary feats, such as searching billions of pages in the blink of an eye. However, that near-instantaneity on the computing dimension is, these days, arguably the easiest to get a handle on — and something we have long since taken for granted. What’s more nuanced is how the search engine appears to evaluate and weigh information.

That’s where web crawlers can screen what motivates: like possibly prioritizing commercial partners, and on occasion seeming to favor particular social and political messages. Or so it seems. Given the stakes in company revenue, those relationships are an understandable approach to running a business. Indeed, it has been reported that some 90% of earnings come from keyword-driven, targeted advertising.

It’s no wonder Google plays up the idea that its engineers are super-smart at what they do. What Google wants us to understand is that its algorithm is complex and constantly changing, for the better. We are allowed to know that when Google decides which search results are most important, pages are ranked by how many other sites link to them — with those sites in turn weighted in importance by their own links.

It’s also obvious that Google performs common-sense concordance searches on the exact text of your query. If you straightforwardly ask, “What is the capital of France?” you will reliably and just as straightforwardly be led to a page saying something like “Paris is the capital of France.” All well and good, and unpretentious, as far as those sorts of one-off queries go.

But what might raise eyebrows among some Google users is the placing of commercial sites above or at least sprinkled amidst factual ones. If you ask, “What do cats eat?” you are led to a cat food manufacturer’s website close to the top of the page, with other informational links surrounding it as if to boost credibility. And if you type “cat” into Google, the links that we recently found near the top of the first page took us not to anything furry and feline  –  but to clunking, great, Caterpillar machinery.

Meanwhile, take a subject that off and on over the last two-plus years has been highly polarizing and politicized — rousing ire, so-called conspiracy theories, and presumptuousness that cleave society across several fronts — like the topical query: “Do covid vaccines have side effects?” Let’s put aside for a moment what you might already be convinced is the answer, either way — whether a full-throated yea or nay.

As a general matter, people might want search engines to reflect the range of context and views — to let searchers ultimately do their own due diligence regarding conflicting opinions. Yet, the all-important first page at Google started, at the time of this particular search, with four sites identified as ads. Followed by several other authoritative links, bunched under ‘More results’, pointing to the vaccine indeed being safe. So, let’s say, you’ll be reassured, but have you been fully informed, to help you understand background and accordingly to make up your own mind?

When we put a similar query to Yahoo!, for comparison, the results were a bit more diverse. Sure, two links were from one of the same sources as Google’s, but a third link was quite a change of pace: a blog suggesting there might be some safety issues, including references to scholarly papers to make sense of the data and conclusions. Might one, in the spirit of avoiding prejudgment, conclude that diversity of information better honours searchers’ agency?

Some people suggest that the technology at Google is rooted in its procedural approach to the science behind it. As a result, it seems that user access to the best information may play second fiddle to mainstream opinion and commercialization, supported, as it has been said, by harvested user data. Yet, isn’t all that the adventurist economic and business model many countries embrace in the name of individual agency and national growth?

Google has been instrumental, of course, in globally democratising access to information in ways undreamt of by history’s cleverest minds. Impressively vast knowledge at the world’s fingertips. But as author Ken Auletta said, “Naïveté and passion make a potent mix; combine the two with power and you have an extraordinary force, one that can effect great change for good or for ill.” Caveat emptor, in other words, despite what one might conclude are good intentions.

Might the savvy technical and business-theoretical minds at Google therefore continue parsing company strategies and search outcomes, as they inventively reshape the search engine’s operational model? And will that continual reinvention help to validate users’ experiences in quests for information intended not only to provide definitive answers but to inform users’ own prioritization and decision-making?

Martin Cohen investigates ‘How Does Google Think’ in his new book, Rethinking Thinking: Problem Solving fro Sun Tzu to Google, which was published by Imprint Academic last month.

Tuesday 26 May 2015

How Google and the NSA are creating a Jealous God

Posted by Pierre-Alain (Perig) Gouanvic




Before PRISM was ever dreamed of, under orders from the Bush White House the NSA was already aiming to “collect it all, sniff it all, know it all, process it all, exploit it all.” During the same period, Google—whose publicly declared corporate mission is to collect and “organize the world’s information and make it universally accessible and useful”was accepting NSA money to the tune of $2 million to provide the agency with search tools for its rapidly accreting hoard of stolen knowledge.
-- Julian Assange, Google Is Not What It Seems

Who is going to process the unthinkable amount of data that's being collected by the NSA and its allies? For now, it seems that the volume of stored data is so enormous that it borders on the absurd.
We know that if someone in the NSA puts a person on notice, his or her record will be retrieved and future actions will be closely monitored (CITIZENFOUR). But who is going to decide who is on notice?

And persons are only significant "threats" if they are related to other persons, to groups, to ideas.

Google, who enjoyed a close proximity with power for the last decade, has now decided to differenciate Good and Bad ideas. Or, in the terms of the New Scientist, truthful content and garbage.
The internet is stuffed with garbage. Anti-vaccination websites make the front page of Google, and fact-free "news" stories spread like wildfire. Google has devised a fix – rank websites according to their truthfulness.
Google's search engine currently uses the number of incoming links to a web page as a proxy for quality, determining where it appears in search results. So pages that many other sites link to are ranked higher. This system has brought us the search engine as we know it today, but the downside is that websites full of misinformation can rise up the rankings, if enough people link to them.
Of course, it is not because vaccine manufacturers are exonerated from liability by the US vaccine court that they are necessarily doing those things that anti-vaccine fanatics say. Italian courts don't judge vaccines the same way as US courts do, but well, that's why we need a more truthful Google, isn't it?

Google will determine what's true using the Knowledge-Based Trust, which in turn will rely on sites "such as Snopes, PolitiFact and FactCheck.org, [...] websites [who] exist and profit directly from debunking anything and everything [and] have been previously exposed as highly partisan."

Wikipedia will all also be part of the adventure.

What is needed by the intelligence community is an understanding of the constellation of threats to power, and those threats might not be the very useful terrorists of 9/11. What is more problematic is those who can lead masses of people to doubt that 19 novice pilots, alone and undisturbed, could fly planes on the World Trade Center on 9/11, or influential people like Robert F. Kennedy who liken USA's vaccine program to mass child abuse.

These idea, and so many other 'garbage' ideas, are the soil on which organized resistance grows. This aggregate of ideas constitutes a powerful, coherent, attractive frame of reference for large, ever expanding, sections of society.

And this is why Google is such an asset to the NSA (and conversely). Google is in charge of arming the NSA with Truth, which, conjoined with power, will create an all-knowing, all-seeing computer-being. Adding private communications to public webpages, Google will identify what's more crucial to 'debunk'. Adding public webpages to private communications, the NSA will be able to connect the personal to the collective.

And this, obviously, will only be possible through artificial intelligence.

Hassabis and his team [of Google's artificial intelligence program  (Deepmind)] are creating opportunities to apply AI to Google services. AI firm is about teaching computers to think like humans, and improved AI could help forge breakthroughs in loads of Google's services [such as truth delivery?]. It could enhance YouTube recommendations for users for example [...].

But it's not just Google product updates that DeepMind's cofounders are thinking about. Worryingly, cofounder Shane Legg thinks the team's advances could be what finishes off the human race. He told the LessWrong blog in an interview: 'Eventually, I think human extinction will probably occur, and technology will likely play a part in this.' He adds that he thinks Artifical Intellgience is the 'No.1 risk for this century'. It's ominous stuff. [ You can read more on that here..]

May


help us.