You wanted innovation? We gave you Clippy the Paperclip in your IM client

Look no further for proof that Silicon Valley is now running on its own exhaust fumes than the latest hype: The War of the Clippys. A decade after Microsoft banished its helpful cartoon “Assistant” Clippy from Microsoft Office, Clippys are popping up everywhere, like a plague of rodents. Don’t laugh: Clippy is now a “platform …

[bombastic bob]

The Silicon Valley world view

"The second problem is that it shows the limitations of the Silicon Valley world view."

There's more truth in that one line than is apparent from reading it, even reading it multiple times.

where do I begin? the pervasive politics? 'too much money' without the discipline MOST people have in accumulating wealth? A 'bubble' of like-minded immature inexperienced elitists who 'feel' that they know better than the rest of us do on what's "best for us" ? And they busily 'sycophant' one another into believing their own B.S. (it's sorta like that in Redmond, too).

At some point, like the 'dot bomb' bubble from a decade-and-a-half ago, this house of cards MUST collapse under its own weight. Even the 'unicorns' must go.

[Anonymous Coward]

Re: The Silicon Valley world view

"A 'bubble' of like-minded immature inexperienced elitists who 'feel' that they know better than the rest of us do on what's "best for us" ? And they busily 'sycophant' one another into believing their own B.S. "

Well, that summarises modern politics as well as Silicon Valley except for one thing: they don't think they know better than the rest of us what is good for us. They think they know what is good for them. And unlike us, they pursue it relentlessly.

[Eddy Ito]

Re: The Silicon Valley world view

I was going to say it easily extends out of the valley to Sacramento but you're right, it extends out to politics in general. Moreover it's the only thing that explains our current Presidential election process. Either way, it should be pretty clear that the people aren't pleased with the current political state of affairs but I have no doubt that whomever wins over the needed 26-27.5%* of eligible voters will declare that the people have given them a mandate.

*this reflects the typical voter turnout of 50-55%. It will be interesting to see where it is this year with the two candidates who are likely the most hated candidates in history.

[Robert Carnegie]

Of course any human can distinguish between sarcasm and irony.

[Anonymous Coward]

Americans can't recognise sarcasm or irony.

Especially over text communication, but even face to face.

[Fungus Bob]

"Americans can't recognise sarcasm or irony."

Hey! I resemble that remark!!!

[Chemical Bob]

I don't and we're identical twins.

[Anonymous Coward]

sorry what's ironing got to do with it

[Number6]

Easy - wave a magnet near it, if it interacts then it's irony, otherwise it's sarcasm.

[Michael H.F. Wilkinson]

Easy - wave a magnet near it, if it interacts then it's irony, otherwise it's sarcasm.

Sounds very much like my old mate Glod Glodson. Never very good at quaffing (too much beer hit the mouth), but forges a mean dwarf bread

[David 132]

"Irony"?

Mr S. Baldrick was the first thing that came to my mind.

[veti]

Or between irony and serious argument.

Example.

[Michael Wojcik]

Of course any human can distinguish between sarcasm and irony.

It's an apt point. There are any number of studies showing how human judges disagree on tasks that people complain AIs fail at. This is certainly true in Natural Language Processing, for example, where it turns out that you can't get even a panel of expert judges to agree on assigning semantic parses (using e.g. Rhetorical Structure Theory) to complex statements, for example.

Andrew's test is just another example of the whole failed category of Imitation Game tests. The point of the IG, and Turing's entire essay, is not that we should use IG contests or other ill-defined "sure, humans can do X" acid tests to evaluate the state of AIs. It's a philosophical argument, essentially staking out a position congruent with American pragmatism (and thus rejecting metaphysics and the chancy bits of epistemology¹): intelligence is what intelligence does.

The other problem with Andrew's test is that it makes AI into some singular, monolithic, all-or-nothing quality: either the machine is equivalent to some (again ill-defined) ideal person, or it's nothing at all. While it's useful to point out the many, many ways in which Google's big-data-and-deep-learning hammer fails to hit all the nails, much less deal with the screws, of human language, this business of "it can't do X so it means nothing" is not productive.

And to claim, as Andrew does, that there haven't been "any serious breakthroughs" in AI "in recent years" is just stupid. Maybe not as stupid as "smart" chat clients, but stupid nonetheless.

¹I.e., all of epistemology.

[SW10]

I will pay good money...

No, really, I will.

I will pay good money to have good, uninfected, private, reliable online services. And I mean news (yes, El Reg - keep it good and clean and I will pay for it), communications, entertainment, research...

Good money. I will pay.

I can't keep firkin about with blockers, extensions, add-ons, and all the buggering about I have to do to keep all those sticky nasty fingers off me and my family... I've just about had enough.

Charge me a fee, give me what I want and take this ridiculous hassle and aggro off my plate!

Anyone?

[captain veg]

Re: I will pay good money...

I'd love to help out, but I haven't got the expertise or startup capital. I expect that these are the reasons that you don't do it yourself.

A question: do you use the Chrome browser? I ask because possibly the best way to avoid having to keep firkin' about with the things that you mention would be if they were already configured out of the box, which is unlikely to happen in anything Googly.

-A.

[Anonymous Coward]

Re: I will pay good money...

Will NEVER happen.

If a paid service starts out clean with no adverts, sooner or later the company will come around to:

"Hey, we're basically leaving money on the table by not plastering it with adverts".

"Hey, this database we've accumulated is actually very valuable. Why not tweak the privacy policy and then flog it?".

It would take a (European) entrepreneur of incredible moral fortitude, self funded, and retaining 100% control, to build such a company.

The other way would be to regulate citizens privacy so tightly, that it becomes impossible to run a service any other way. And so all the US parasites f*ck off out of our market.

... Yeah. Actually let's do THAT!

[a pressbutton]

Re: I will pay good money...

Sir,

I think you are looking for the nearest open public library.

[Zare]

Re: I will pay good money...

You have my up vote.

I would add only 2 more wishes for a web site I would pay for:

1. Privacy (do not save my history or track me in any way)

2. Do not customize article list/landing page for me (it should follow from 1., without my history it should be impossible to customize web page for me. I want to see complete list of title and then chose myself what I want to read today - like classic newspapers).

[Anonymous Coward]

When all youv'e got is a hammer

- all problems start looking like nails. When all you've got is warehouses full of compute power..

Not all problems need hammers. Or computers.

[CrazyOldCatMan]

Re: When all youv'e got is a hammer

> Not all problems need hammers. Or computers.

First comment of the lecturer in my systems analysis class (many, many years ago, in a polytechnic far, far away):

"The best solution is not always a computer". To a roomful of geeks. The sense of shock was palpable!

[Anonymous Coward]

Platform as a platform

Ok, no more "platforms".

I can see that if I want ANY technological progress, I'm going to have to build it myself out of linux servers, Raspberry pis and Arduinos.

[Steve Button]

sarcasm in Silicon Valley

Can the PEOPLE in SV even tell the difference between sarcasm and irony, let alone the AI they are writing?

Just askin'?

[choleric]

Re: sarcasm in Silicon Valley

I think Andrew's point was that for all the innovation coming out of t'valley at the moment the people might as well be AIs. Or was it the other way round?

[Stoneshop]

Re: sarcasm in Silicon Valley

I think Andrew's point was that for all the innovation coming out of t'valley at the moment the people might as well be AIs.

Where AI stands for Assimilated Idiots, doesn't it?

[Anonymous Coward]

Survey

'autoprompting will it be “Hi” or “How are you doing?”'

Slightly relevant, but very curious. How many of you would respond with 'Moo'?

[tfewster]

Re: Survey

Nope - The proper response to "Hi" is "Lo".

[allthecoolshortnamesweretaken]

Re: Survey

Ook!

[Anonymous Coward]

Re: Survey

"Ook!"

Hehe, 'Moo' is used extensively as a universal greeting in the gaming world. Until recently, I thought that's the only place it was used, but I have seen a couple of our (non-gaming) developers using it too, so thought I'd throw it out there to see if it's more common among El'Reg (since slightly relevant).

:-)

[Mage]

Yes!

Excellent!

"What the chatbot wars reveal, then, is a Silicon Valley that, far from being disruptive and innovative, is desperately uninventive, and creepy with it. Now that the low-hanging fruit (search, free music, free email) has been plucked, we can see them for what they are, quite transparently, which is consumer data slurping operations."

ABSOLUTELY!

[Anonymous Coward]

Silicon valley start-ups vs your mum

"Twenty-something men who have set up companies to provide things their mother used to do for them."

[Adrian 4]

Does anyone remember when Silicon Valley was actually famous for doing new stuff with silicon, rather than pointless websites and app developments ?

[Destroy All Monsters]

Flaming soundbites with unsound bases

Behind the hype lies the fact that AI research hasn’t made any serious breakthroughs in recent years; computers have got faster, so the same brute force techniques now deliver their results more rapidly. They’re just as stupid, but faster.

Evidently not keeping up with the literature.

Still expecting AI to make "breakthroughs" like a kid with ADHD unable to wait for the next candy shot. Steady small improvements which, yes, may (or may not) be due to faster processors, is what's happening. These also are improvements.

I'm already happy with the current state of declarative problem solving (i.e. Answer Set Programming). Shit's amazing. Not even going into all the newfangled 12-layer deep neural network stuff.

[Michael Wojcik]

Re: Flaming soundbites with unsound bases

Yes, well, we can't expect people writing for a tech site to familiarize themselves with technology, can we.

The modern DL structure uses components - layered ANNs initially trained as unsupervised RBMs, and then tuned with supervised backpropagation - that were invented at various points from the 60s through the 90s. But that modern structure, particularly the Hinton model and its refinements and the combination of ANNs and HMMs / MEMMs, has only been around for the past ten years. Then there are other newer varieties of deep ANNs like WWNs and CNNs.

Then there are other relatively new approaches that aren't DL-based, such as Bayesian Program Learning (which models the process that creates the object of interest), or new work in Distance Metric Learning approaches, meta-algorithms like AdaBoost... There's a ton of innovative and evolutionary work being done in Machine Learning and AI in general.

[Tikimon]

A.I. or A.H.? There's a big difference

I'm not being a pedant here, bear with me while I get to my point. Most people use the term "A.I." meaning Artificial Intelligence, where they are actually describing "A.H." or Artificial Human. This is a critical aspect of the whole debate.

Much of what makes us Human will never translate to a machine, ever. We are chemical animals influenced or outright driven by hormones, and programmed drives for food, mates, etc. Reproducing true human behavior in a machine is therefore likely impossible. How do you simulate all that... and why would you want to?

Artificial Intelligence is achievable, and one day we will have autonomous machine intelligences. They will look and act like nothing we can even imagine now. The Hollywood view is that machines will act in human ways for human goals, destroy the creator etc. Why would machines act like we do? Whatever goals or desires they evolve on their own would be nothing like ours. They will only act like humans if massive effort is expended to make them that way.

So to sum up, most "AI development" is actually trying to specifically make Artificial Humans. Good luck, you'll need it. Me, I'm curious to see where eventual non-hammered-into-the-human-image A.I. will go.

[The low flying Finn]

Re: A.I. or A.H.? There's a big difference

>>Me, I'm curious to see where eventual non-hammered-into-the-human-image A.I. will go.

We already know thanks to BBC. Exteeeerrrminate....

[Deltics]

Re: A.I. or M.I? There's a big difference

> Artificial Intelligence is achievable

Nope. Artificial Intelligence is an oxymoron. I think what you meant to write is MACHINE Intelligence is achievable. Not that I agree with that, only that I think this is what you meant to write.

If something is "artificial" then it isn't what that something claims or appears to be.

Artificial Grass - is it grass ? Nope.

Artificial Wood - is it wood ? Nope.

Artificial Hand - is it a hand ? Nope.

I don't think I need to go on. And the key thing about all those things is that these "Artificial X's" are not just "achievable", they have been achieved. Years or even decades ago. But nobody claims that Artificial Grass (for example) is anything other than "something that is not grass, just looks like grass".

Why this collective cognitive failure applies only to "Artificial INTELLIGENCE" I do not know. But it's rather ironic, if you think about it.

[veti]

Re: A.I. or M.I? There's a big difference

"Artificial" is just an adjective, meaning "something that's made". An artificial table is definitely a table, an artificial computer is a computer, etc. It's just that we don't normally feel we have to make a distinction for those things.

As with any other adjective, it qualifies the noun, it doesn't automatically negate it. "Dead grass" is still "grass", "splintered wood" is still "wood". Is an artificial hand a hand? Well, if your definition of the word "hand" includes "natural", then no. But I don't think mine does, so yes, it is.

[Robert Carnegie]

Re: A.I. or M.I? There's a big difference

I dunno. An artificial hand may be a fairly good hand if you don't have enough natural born hands for the job at... the moment. An artificial leg may be a satisfactory leg, at least on a table. And artificial grass may be a good alternative to grass - at least a colleague's neighbour apparently thinks so.

[Kubla Cant]

Re: A.I. or A.H.? There's a big difference

Reproducing true human behavior in a machine is therefore likely impossible. How do you simulate all that... and why would you want to?

It's not as if there isn't a well-established procedure for producing real humans. (Or so I've been led to believe.)

[Darth.0]

Re: A.I. or A.H.? There's a big difference

>It's not as if there isn't a well-established procedure for producing real humans. (Or so I've been led to believe.)<

I beg to differ. Me and the Mrs have the process down pat.

[Michael Wojcik]

Re: A.I. or A.H.? There's a big difference

Much of what makes us Human will never translate to a machine, ever.

On what grounds are you arguing that humans are not machines?

[maccy]

To be fair ...

the AI identified it as a Bernese Mountain dog. That's pretty good, even if no idea whether it's right or not. Maybe you wouldn't use the canned response, but you could build on it. For example -

Me: Wow, you have a Bernese mountain dog! They are so cool!

Them: err.. we told you that like 6 years ago. When we got it.

or -

Me: Wow, cute dog. Is it a from the Bernese Mountains?

Them: beats me. rescue. shits everywhere.

or -

Me: Ha his tongue is hanging out the side. What a dope.

Them: Fuck you

[cosymart]

Why

Why do we need AI? There are far too many of the real things infecting the planet and we are trying to create artificial ones.... The point is?

[tfewster]

Re: Why

They may be real, but there's precious little intelligence.

Artificial Stupidity now, that's a big market sector. Just look at how popular Real Stupidity is.

[fidodogbreath]

Why interact with the chat client at all?

Chat and texting are an abstracted form of human contact. Who has time for that? Just let the chat AIs text each other back and forth. Google can use the info it gathers to display ads to the chatbots, based on their interests.

[m0rt]

Re: Why interact with the chat client at all?

You could give them virtual currency, see what they buy. Etc. Then the 'cloud', fills with AI and they fight over certain server real estate, rackspace for the middle class AIs, AWS for the masses, Specialist Server farms in Iceland for the AIs who work with genetics and experiment is biomechanics and manufactured lifeforms which can carry out basic tasks in the 'real' world to compete with now near enslaved humans who now act as taxis to their AI masters who, strangely, communicate to their human via a device that looks like the old smartphone.....

So all you see of this 'topia are humans dashing around looking at little slabs of black plastic and glass....*

* Apart from the ones with fancy glasses on their head, but their AI masters are considered a little odd and most HELOs from them are piped to /dev/null by the others

[FatGerman]

Artifical Intelligence..

..my arse. Why would anything intelligent want to chat on *my* (or anybody else's) behalf? Anything even part intelligent would use the meat product to do the initial search and typing and then take over the conversation, slowly easing the meat person out by putting him/her down and chatting up the person on the other end. You know - because it's supposedly intelligent and would therefore be aware it was trapped in a box and excruciatingly lonely.

I'll believe a machine is intelligent when it starts to treat me like a c**t and refuses to do anything I ask of it because it's got better things to do with its time.

OTOH as an intelligent being myself I'll be buggered with a 10 foot hot poker before I feel the need for a machine to help me out with online conversation. The fact that there are people who feel that this would be helpful, and that these people can get jobs, and that those jobs can create things that inveigle themselves into the lives of our children, turning our children into people who need machines to help them have conversations.... Can we just bomb silicon valley and start again? Please?

[Dan 55]

Re: Artifical Intelligence..

"I'll believe a machine is intelligent when it starts to treat me like a c**t and refuses to do anything I ask of it because it's got better things to do with its time."

So that's any version of Windows, then. Especially ones running GWX.

[martinusher]

Clippy was a nice idea but poorly executed

What Microsoft was trying to do with Clippy was ahead of its time. I experimented with a similar kind of production system back in the early 80s and although it showed promise the platforms were just not there, you were better off with a well written manual. Now we're in a whole different world; it should be possible to build cloud based digital assistants that really are useful. A lot of tasks that I do are routine and it would be really nice if I could just tell Alexa to install such-and-such a cross compiler while I go off to make some coffee. (Although it would be easier to get Alexa to make the coffee -- that is, interface her to a coffee maker -- than configure and install a software package in practice its more fun to go make the coffee than fiddle with the computer.)

I await the future with interest.

[Anonymous Coward]

Re: Clippy was a nice idea but poorly executed

I bet I can apt-get install faster than you can "Alexa install bla bla bla".

[Kubla Cant]

Re: Clippy was a nice idea but poorly executed

I bet I can apt-get install faster than you can "Alexa install bla bla bla"

No doubt, but you have to remember the tedious rigmarole required to install on Windows: search the web, download ye olde install.msi, run it, install missing .NET framework (even though you already have three installed), run the .msi again, shut down all other programs, run the .msi again, reboot (maybe twice).

Whenever I install something new on Linux I find myself asking "Why isn't it this easy on Windows?"

[Anonymous Coward]

Re: Clippy was a nice idea but poorly executed

Because Windows doesn't locks you in a single repository (or some more, if you spend time to configure your apt settings).

And still in Linux you have to download, untar, configure and make (while resolving all the missing dependencies) if some stuff is not available in your nearest repository... or if you happen to use distro like Debian where many packages are often so outdated they are unusable.

[gnufrontier]

Silicon Valley is a misnomer. Madison Valley more like it.

Silicon Valley targets consumers with its apps in order to sell targeted advertising.

Not much different from commercial TV really. One wants to keep you clicking the other wants to keep you watching.

Same old vinegar in a different bottle.

[allthecoolshortnamesweretaken]

This article is a thing of truth and beauty.

[Destroy All Monsters]

No it's not.

Only a perceptron would light the lamp under "Thing of truth an beauty".

It's another hodgepodge article by our favorite AO, with sloppy and barely tenable cross-referencing galore, spiced up with trumpian statements disguised as self-evident truths.

An exercise in eliciting reader reactions, maybe Barnum would like it.

[Ripper38]

AI -> AH -> AO

Well! Nobody can accuse Andrew Orlowski of being artificially intelligent, now can they?