Unsupervised Computing, Chapter 1: Search

Now that we roughly defined the distinction between supervised and unsupervised computing, we can take our first step:

Opt out from Google search to a search engine that does not track you and exploit your behavioural data.

Only inspired action carries meaning. So what is the inspiration here? Why would we want to opt out of Google search?

• Google shows you several Ads before you get to the actual organic search results.
• All your searches are tracked — and what you search for says a lot about you. It shows google what your interests are and what you care about. This information is very valuable to advertisers.
• Let’s say you search for “pregnancy test” — it will make sense then to show you either ads for baby cloths or for contraceptives. Your later searches will tell Google exactly what route to go.
• Look at your own search history and go a couple weeks back. Do you remember? Google does too.

There are many alternatives to Google, but if your concern is not being tracked, try DuckDuckGo. And here we go with the tradeoff:

Because Google tracked your every move, you are used to “personalised” search results. Google’s algorithms are really good at predicting what results matter most to you — because Google has collected data about you, probably for many years.

So the results of any other search engine will of course not be as perfect as Google’s. Especially if that search engine simply refuses to track people.

The good part is: in most cases that is hardly noticable. 80% of my search works perfectly fine in DuckDuckGo. For some special stuff I need to fall back on Google, especially local results: since Google has the immense backbone of Google Maps and since all businesses and other locations want to be findable there, all that data is reflected in searches for local results.

Falling back on Google where DuckDuckGo fails is really simple to do though: there nice little feature called “bang search”. With certain key letters you can forward your search to, say Google.

• !g xxxxy will lead you to a Google search on xxxxxy
• !yt xxxxy will lead you to a YouTube search on xxxxy
• !w xxxxy will lead you to a Wikipedia search on xxxxxy

…and so forth, you get the idea: https://duckduckgo.com/bang

However, for now, let’s commit to trying it out. How is this:

For two weeks I will keep DuckDuckGo as my default search engine. I will only fall back to Google if I cannot find what I need.

(If you do not know how to change your “default search engine” — don’t google it… rather search for it on DuckDuckGo.)

Two weeks should give you a) a solid feel about DuckDuckGo and b) get you over the initial, very human hurdles of “this is new I don’t like it” and “this is not as perfect as it was before and I don’t like it”. It’s a tradeoff. You give up a bit of convenience. You gain back a bit of privacy and autonomy.

If DuckDuckGo is now your default search engine: Congratulations! You just did your first step in unsupervised computing.

And since we all love some badges and stickers, I might make someday… not now though.

In the next chapter we will up our game a little — and change our browser 🙀

Until then, stay safe, stay armed ✌️

People pay 20 bucks for a notebook, then write in it with a 20 cent pen they found in a goodie bag.

Unsupervised Computing, Introduction: What is unsupervised computing?

To answer that, let’s first look at what we might call “supervised” computing.

In supervised computing you trade autonomy and ownership for safety and convenience. That might sound bad at first, depending where you’re coming from. It is actually a pretty good deal for most people — at least in the short term.

Some examples:

• I have no idea what an url is or what exactly “a browser” is
• I’m not aware that my ISP and mobile provider know what URLs I surf to or who each of my devices talk to and when
• I’m not aware that email is much more like a postcard and not a letter at all
• I consume and communicate mostly via big tech apps like Whatsapp, Instagram, Amazon, Netflix, Facebook, Youtube, Twitter, Google, Apple and so forth
• I’m given access to content based on licenses and subscriptions rather than actually owning anything
• I’m unaware of how deep DRM is abused by corporations to force me into buying more disposable shit. I have not idea that circumventing superficial DRM to repair broken things is a crime that land me in jail.
• I don’t even know that there is a plethora of apps outside the walled gardens of Apple’s App Store and Google’s Play Store.
• I have zero idea of what is behind all those shiny buttons. I have never heard of a “terminal”.
• I cannot open, understand or even repair my devices. Opening them usually breaks them.
• 12 years after the invention of Bitcoin and valued close to 20k a piece, all I know about it is still “people use Bitcoin to buy drugs and guns”

In supervised computing you trade autonomy and ownership for safety and convenience.

Earlier I said that is actually a pretty good short-term deal for many people.

And it is a terrific deal for big tech. If you’re not aware of how “big data” works, if you do not know how Facebook ads work, if you’ve not heard of “surveillance capitalism” — I’d advice for some reading. Please don’t “google” it though. Use another search engine.

Now we can finally answer what unsupervised computing is:

In unsupervised computing you trade in some of the safety and convenience big tech tech provides — and in exchange you get back some autonomy and ownership.

It’s really that simple.

It’s like biking without training wheels. It’s like letting go off the rail on a large staircase. It’s like swimming in water that is deeper than your height.

For some, the smell of autonomy and a tiny bit of risk is what makes life fun. They start mountain biking or bouldering or boxing.

Others prefer watching mountain biking on TV. And that is fine too.

Unsupervised computing is not a virtue. You’re not better if you do unsupervised computing. Especially since the first steps are incredible easy and simple to do.

In the next chapter, we will look into a couple of those easy and simple quickwins in order to gain back some ownership, autonomy and privacy.

Until then, stay safe, stay armed ✌️

#unsupervisedcomputing

A colleague of mine give a small book to people when they stop working at a company. That is lovely.

I wonder though: why not give it around when you start?

Nagging bait

Dear Kids.

Here’s something my dad taught me when I was about 16. At the time he was doing print stuffs, like paper bags for bakeries, catalogs and the like. So he was getting feedback on graphic design work from people who have no fucking clue what they were talking about. Plus: they were clients. Clients always need something to critisize. They need to be pampered and appreciated for their very valuable domain knowledge and general smartness, and they also need to feel you doing as they say since they pay you.

If you do everything right, if you give a perfect solution and deliver 120% and also early — the client will find something. They have to. Otherwise they would feel useless I guess.

So my dad would build in “nagging bait”. Nagging bait is a very obvious “mistake”, something everybody can spot on first sight and be absolutely right about nagging, correcting and critisizing. He would then expand on the feedback, ask more questions and make sure the pampering came through. Then he would “iterate on the design”, which meant waiting for a couple days, then present the original finished piece before the nagging bait was built into it.

It might feel foul at first. But most clients of design work are themselves foul in the first place. They wouldn’t tell a plumber how to fit the pipes of their sink. But sure as hell will they tell you how the typeface isn’t quite what they like.

Hopefully needless to say: unless you do very thorough, thoughtful and high-quality work, don’t even think about using nagging bait. Unless you have at least two reasons for every design decision you made, stay off political games like that. You are probably producing enough real mistakes as it is.

Nagging bait is especially useful when you try to do something for a comittee. Comittees are the worst. They are usually unaware of how political they behave and how difficult it is to get some decisions out of them. On nagging bait they can instantly agree. They can chew it from left to right and back again and then discuss it forwards and backwards and feel like they just had a very productive meeting.

It doesn’t just work for design work btw. You can put nagging bait into everything you create. As long as you do it with the right intentions, it‘s allright.