DK_en 1x06 - Lenin in Silicon valley

Episode first aired on June 1, 2018. Listen to the audio on Spreaker.com

It's been quite a while since my last English installment.

Bit of laziness, yes, also a new job, and a quite demanding one, so I had to choose and I chose to put the English edition on hold.

Now that the GDPR is a reality, it's time to revive this podcast in English.

Today's installment is not about Facebook, there's already too much talk about it, but about another one of the GAFAMs. The one that is less easy to hate, because it's not represented by an entitled nerd on a ego trip, it simply behaves like an entitled nerd on a ego trip.

I'm talking about Google.

The first thing I'll mention is Project Maven, a piece of news from a couple months ago.

What is Project Maven? It's a mega contract, Google has signed the US Department of Defence where Google would put its AI expertise at work on military recon.

Google's AI's will analyse drone video fields and identify potential targets, taking drone assassinations to an industrial scale.

Beyond any consideration on the operational limits of what passes for AI these days, and God knows if there are some, the idea is so repulsive that a few hundred of Google's own engineers have filed their resignation in protest.

Since this is already old news, I will only say this.

Project Maven will make the US look so bad that in comparison Vietnam, Iraq and Afghanistan will look like something to be proud about.

More recent news is that of Google Duplex, where Google solves the age-old problem of, brace yourselves, making a phone reservation. I read the rationale as was given by Duplex's Product Manager. It was basically three points.

One, our mission is to put clients in touch with businesses. Two, about 80% of small businesses do not have an automated reservation system. And three, so we created a vocal bot to make phone reservations.

If the logic escapes you, worry not, this is classic example of ad hoc reasoning dates back to the time of Aristoteles.

Today this kind of bogus inference is known as "A-Therefore-B", and is most common among, you know, salespeople, marketers, and of course politicians.

Sadly though, it works. Years of dumbification of the public are finally paying off.

Anyway, Google has actually built the vocal bot and dubbed it Duplex.

At the latest Google I/O, one of the annual groupie meetups that passed for industry conventions, Duplex has been met with Marvel and Enthusiasm, onetheless, and this from the industry press, a prime example of the level of scrutiny that defaults tech companies.

Marvel because a bot showed, you know, human level dialogue fluency, not exactly an earth shattering result, given what we're already accustomed to.

And enthusiasm, because the bot used cool sounding utterances like, er, or the inevitable, like, as in:

I'm looking for, like, Thursday morning around, er, like, 11.

That must be the oldest trick in the box. Even Eliza could do something like this, back in the 70s.

A rare unintoxicated observer noted,

We thought robots would speak like Hull in 2001 a space odyssey, instead they speak like millennials.

He didn't exactly mean that as a compliment. But let's stick to the point. What's the use of a product like Duplex?

Clearly, it has been shown it will make a phone reservation for you. Now that's a problem.

Let's take the demo. It's on YouTube. You can watch it when you want. So let's take that at face value.

Let's assume that the voice production was actually in real time, and there was no editing.

Duplex is not the usual solution to nobody's problem that's a Silicon Valley staple. There are serious issues here. The first being, the bot is apparently indistinguishable from a human.

This is actually marketed as an achievement. But this is a Pandora's box, not just because decent design demands, that an artifact tell me what it is, how it works, and how it is working.

Perspicuity, the ability of an artifact to communicate its abilities, its function, is fundamental.

Well, Duplex does not care about being perspicuous. It pretends and pretends very well to have dialogue and comprehension abilities comparable to those of a human, which means, its drawbacks and limitations will only be discovered when problems arise, that means, at the wrong moment.

I'll explain. If I'm before a machine of sorts, I expect to have the limits of a machine. But if I am before a human being, I expect actual human behavior.

The problem with something that pretends to be something else is this: things get used for what they suggest. Design professionals call it allowance.

So, the problem with having something that sounds like a human being is that people will expect it to react like a proper human. Except of course, that when there is a problem, and there will be problems, there always are, people will realize they've been speaking to a stupid machine.

This is the road to disaster. Now, if the use case for Duplex is really to reserve a haircut, of course, we reserve extinction.

But supposing there were less damning use cases, what would they be? I for one see a market in porn hotlines, but apart from this, I see very little.

Silicon Valley solutionism loves to mention efficiency.

But as I've been repeating for decades now, efficiency is okay if you are already sure you're doing the right thing. Otherwise, your goal really should be effectiveness.

Now, we can assume that Google Duplex makes it more efficient to make phone calls like "reserve X from shop Y for such day at such time". That is assuming Duplex works as advertised.

It's a big assumption. Now, are we sure this is the right thing to automate? How many reservations are we supposed to make in a day? Second, imagine you are on the receiving end of one such call. Now, that you recognise the bot as a bot or not is unimportant.

At the slightest problem though, the bot will behave like a bot anyway, which means if the reservation it requires or the alternatives it can handle is unavailable, you lose a client. While with a human call instead, you would almost always find an agreement.

So, let us suppose these things do become widespread.

As a provider of goods or services, what is your best option? Your best option is to switch to automated responders so that you can automate the whole process. But what we are seeing is Google is wanting you to use a mile high tower of software, cloud-based of course, on Google Cloud of course, to make a reservation system, you could have working today on your standard website.

And we are only talking of innocent legitimate use. But at a technology like this, seems conceived, I say seems instead of is because as a European I respect the epistemological value of doubt.

Seems conceived, I said, with the worst in mind. A technology like this enables, just to say one thing, phone scams at a scale. And as I said before, will do wonders for porn hotlines.

Oh, then of course, it will allow for price dumping on manpower costs in a sector that's already largely commoditized. In plain words, it will create slaves.

Now, are we sure that lowering operating costs in outgoing call centers is a worthy goal to pursue? Put yourselves in the shoes of call center worker, competing with a machine for her job. Can you imagine anything more dehumanizing?

And don't tell me, please don't tell me these jobs are obsolete. They will only be obsolete when no one has such jobs as the only alternative to poverty. So again, are we sure this automation of low-skill jobs is what we need as a society?

Of course, in a perfect future, everybody will have a PhD. But we do not leave in that perfect future and judging from education costs and job statistics, we are nowhere near beginning to prepare it.

Silicon Valley Bros keep repeating they want to change the world. Looking back to the past 20 years, I think it's high time is started asking to change how.

And we've come to the Selfish Ledger. The Selfish Ledger is a project, or better, a concept that reveals the true nature of Google, that of the Church of the Algojerks.

A few basic facts.

Selfish ledger is a homage to Richard Dawkins' book The Selfish Gene.

And the name of an internal video from X, formerly Google X, that's Google's division for wet dreams, marketed under the label of "advanced" projects.

This concept regards turning our smartphones into a global, intergenerational record of personal data of everyone, everywhere. Guess who gets the store and process all these data? Right, that would be Google. And why would it do all this?

But of course, to help us achieve our, and I quote, "life goals". So in the concept, you choose one life goal and the Selfish Ledger starts telling you whether your actions take you closer to or farther from that goal.

Let's make an example. You may decide your "life goal" would be to lose 10 kilograms. And whenever you enter an ice cream shop or pizzeria, the phone goes beep and notifies you that ice cream and pizza are not along the road to a slimmer you.

It's worth repeating, this comes from the Google division in charge of the long haul, in charge of shaping the future. I call this a monument to intellectual misery.

Life as seen through the eyes of a how-to book author. Intellectual garbage to fill the intellectual void that culture has become. And of course, it does not stop here. There's of course lampooning material for years, but so far we've only dealt with the Ledger part. The Selfish part is the more problematic one, and is where Silicon Valley hipsters get all red in the face and start breathing heavily.

The Selfish idea is that the ledger does not need to be limited to a one single life, but could easily be extended and refined over generations. Until at some point, Google's AIs think they know enough of us to decide on their own how to nudge us towards our life goals and also nudge us towards another goal: nothing less than the greater good, or better, its version as selflessly interpreted by your friendly mega corporation, Google.

For instance, the video shows how at some point the ledger finds out it lacks information on the user weight, and therefore decides to 3D print a custom electronics scale designed to respond the personal sense of aesthetics of the user.

I guess it would be too trivial to go on Amazon and buy a Chinese one, plus Amazon is the competition, and apparently the greater good is best pursued by, you know, locking out the competition.

It doesn't stop here. As we said, once the ledger is able to span across generation, limiting its scope to the goals of an individual life seems puny.

Google thinks the ledger should also pursue other goals, bigger ones, to, and I quote again, "make a better world." So 27 years after the dissolution of the USSR, the worst of California's neoliberalism rediscovers the idea of a planned economy and of a technocratic elite whose role is to guide the masses towards its own idea of progress.

Dear United States of America, I hope you understand this is Leninism.

And while we're at it, I think it was Marx once said it history repeats itself first as tragedy, then as farce.

Well, at least the first time this was a proper political theory.
At least the first time was the effing first time. Now it's just hipster bullshit.

Anyway, the result for this is a future as Cyberpunk imagined it over 30 years ago.

Needlessly hyper technological, culturally empty, humanly desolate, and of course, long overdue because after 30 years one must be blind, or a rich white man from California, not to see the limits and the cluelessness of that kind of future. A future where, of course, Google would own the world's data and rule over the "goals to make the world a better place".

I will not waste your time telling you this is exactly the same vision that the Zuck has for Facebook.

These guys made it rich with their two penny ideas and now want to turn their economic power into political power, like any other Plutocrat in history.

Italy's recent history and the US current history teach us that when these people get in power, they show very few, if any, skills and interests beyond their own self-interest.

Next time someone speaks with disdain in China and its experiments in technocratic social control, remember that Google, Facebook, Amazon, Apple, and Microsoft are trying to build the exact same future only with Hollywood-grade aesthetics and storytelling.

Don't you think this looks terribly, sadly, desolately old? I mean, here we are, walking around with supercomputers in our pockets, and all these people can think of are the dreams of a teenager and a future that's one century old?

Let us say it now, before they show up again in a few years, with their perfectly timed confessions, ready to let us know again what is is that we really want:

the creative force of Silicon Valley is exhausted.

For one thing, we can no longer expect innovation from what are now traditional companies, the GAFAMs.

But more importantly, the political project they marketed as innovation has failed, a project that used information technology to pursue occult social manipulation.

Google finally reveals itself for the cult it really is, and more precisely the Church of the Algojerks, the new instrument of colonial domination of the plutocracy formerly known as the United States of America.

I guess this seriously is a step forward from the pretence of revolution and progress we've been fed in the past 40 years. They may produce a few more cool gadgets, but their technical onanism is starting to reek, not of rotten, maybe, surely of old. We now know, once and for all, that the future of society, the future of rights, the future of culture is elsewhere.

In this context, it is only fitting to discover that Google has officially removed its "don't be evil" slogan from its code of conduct.

The candid, if terribly naive, motto, has been replaced by a frankly useless "do the right thing". In case you're wondering, they get to decide what right is, they don't even need a Central Committee or a Politburo.

The late comrade Vladimir Ilich Ulyanov, but will name Lenin, would be proud. I think they now consider it okay for them to be evil. Well, it is surely okay for us to treat them as such. We may learn from the people of Kreuzberg, the Berlin neighborhood protesting the planned opening of a huge Google research center.

We can say with them, "Fuck off, Google."