Tag: education

AI learning to explain itself

Recently, there was news again that indeed, ABC (or The ‘Alphabet’ Company Formerly Know As Google) was developing AI that could improve itself.
O-kay… No sweat, or sweat, qua bleak future for humankind, but …
Can the ‘AI’ thus developing itself, maybe be turned first to learn how to explain itself ..? Then, this [incl link therein!] will revert to the auditors’ original of second opinions … Since the self-explanatory part may very well be the most difficult part of ‘intelligence’, benefitting the most from the ( AI improving itself )2 part or what?

And:

[Improving yourself as the imperative; Frank Lloyd Wright’s Beth Sholom at Elkins Park, PA]

Simply laborious

After some post recently, I was triggered to summarise – and expand …
Since there is a bit of history missing. Being the Theory of Firm. Which is, among other stuff but au fond, about the creation of the manager. As the go-between of ’employees’, as the go-between between the workforce and Capital. As the foreman, the one in charge of coordination – when the people come together because they can achieve more in cooperation than the sum of their individual efforts, through specialisation of labour, those specialised contributions need coming together in one way or another, and the provider of capital (the thing that other raw materials are paid with hence the about-only thing to receive deferred payment; don’t get me started on the so absolute quod-non of the ‘inherent right’ to rent above trivial liquidity and risk compensation…) will want to talk to just one. Originally, sometimes capital provider and leader/cooperation-initiator/manager were one, until external capital was required from parties that extorted control. Big sic there.
Well, now don’t go blaming me that it is on the capital provider side that criminally biased rules and regulations have crept in. Flash capital, extortionist locust ‘capital management’ groups, et al., have forced their mob ways into ‘normal’ conduct — almost; the Rheinland model [first, learn to pronounce that correctly, then, get to understand it fully, then, return and prostate and happily receive your life sentences for your transgressions] still holds sway, and sometimes veers back a bit. A bit.

So yes, to the latter, that is criminal, and the cause of cushions called management layers, ever more wrongly devised and developed. And yes, we would need some totalitarian revision of organisational structures to cure it all. Including, starting with, the redefinition [i.e., throwing away all that function there smoothly now, as they denounce their incapacity to really do what’s really required] of middle management and refilling the positions. Also capping CEO pay to, say, something like 10 times the average pay of the bottom 10% of the workforce. All contribute, and a CEO may need a little compensation for when [not if; should be law..!] something goes wrong, his (sic) head rolls. But not too much. When the CEO is sacrificed, something went so wrong that many will get hit; the CEO being in the best position to survive it, qua social and economic strata he should be in. Workers, much less so; much less opportunity to have built buffers, capice?

But maybe not absolute only Owners and Professionals. That will simply not work. Both sides would, even in an ideal world of perfect information everywhere, be buried under control information to the mountainous levels that they wouldn’t have time left [if you’d need more than 25 hours a day to do your work, just work that little longer!] to do their primary jobs…
But a revisit of Galbraith’ four information processing capacity increasing routes, as here, is desperately necessary… Surely, herein lies the way forward to much better organisational design, integrating the latest of possibilities qua information processing and internal and external networking, making possible the creation of true networked organisations and individuals…?

Oh, and:
[Completely undoctored, also unsmoked, pic from Toronto]

Surge ethiconomics

There was already quite some debate about surge pricing, in particular re [illegal] taxi services.
What I missed so far, are discussions about the economic or raher ethical character of abusing surges and their price tag instabilities. Like, how would you depict such developments in price elasticity graphs; shooting up and down on-curve, and curve shifts included. Is orderly society permissive of such hog cycle disruptions ..? [Term pointing at the characterisation of the CEOs that not want to see anything in/human in what they do]
The asymmetry (shooting) on the curve, is market imperfection; the curve shifts in the long run, are better captured by ‘classical’ economics. Again: the ethical ramifications aren’t value-free (tauto), aren’t of uninterest to anyone that values freedom — as that requires markets to function, which is done by regulating them. The latter is proven so many times I don’t even want to discuss it here.

Stock markets, and stocks, are capped qua max change (volatility spikes), the most extreme competitive markets out there;
why wouldn’t other markets have the same ..?

Your contributions in comments, please… Plus:

[Stable, safe, cleared for use; Madrid]

Visual on socmed shallowness

When considering the senses, is it not that Visual (having come to ther play rather late in evolution or has it but that’s beside the point, is it?) has been tuned to ultrahigh-speed 2D/3D input processing ..? Like, light waves particles who are you fooling? happen to be the fastest thing around, qua practical human-scale environmental signals – so far, yeah, yeah… – and have been specialised to be used for detection of danger all around, even qua motion at really high pace (despite the 24-fps frame blinking).
Thus the question arises: What sense would you select, when focusing on shallow processing of the high-speed response type? Visual, indeed. Biologically making it less useful for deep thought and connection, etc.

Now that the world has turned so Visual (socmed with its intelligence-squashing filters, etc.; AR/VR going in the same direction of course), how could we expect anything else that the Shallows ..? Will we not destroy by negative, non-re inforcement, human intelligence and have only consumers left at the will of ANI/ASI ..?

Not that I have the antidote… Or it would be to Read, and Study (with sparse use of visual, like not needing sound bite sentences but some more structured texts), and do deep, very deep thinking without external inference.
But still… Plus:

[An ecosystem that lives off nanosecond trading – no need for human involvement so they’re cut out brutally; NY]

Per vertical lines of defense

What if … Lines of Defense aren’t three (or four or five) ‘horizontally’, but vertical, like actual protection against things getting out of bounds ..?
Wouldn’t that return the whole concept of 3LD, TLD, Three LoD or what’s your favourite abbreviation, to the already tried and tested process control models of yesteryear and when not if Yes, wouldn’t you be found out to be a sort of bumbling eager beaver when you think you’re still doing great and are Really Important and a GRC star and don’t see your kindergarten Importance is called out to hang high ..?

Because then, you’ll need no more big Risk departments with all the procedural justice, compliance on paper (and actual (operating) effectiveness nowhere!), etc., just some nimble support structure. Then, a major part of the conzulting industry would collapse and core management capabilities would have to be returned to formal and practical education and experience-training.

Oh well, one can dream, can’t one?
And:

[A lot of science and engineering there, inside and out, and how beautiful it is (for it); Valencia]

Profiling the politics of the GDPR

When looking up the definition of ‘politics’, no-one can escape the notion that it regards something-choice or in any form the application of power to make decisions applying to all members of a group.
When looking up what leeways for profiling there is in the GDPR, even when so completely fellow-traveller-like as e.g., here [apart from the many, many more errors of logical reasoning, of thought, and of morality and ethics in that piece], the special category of data immediately springs to mind … that is about political opinion – representing the individuals’ autonomy in matters of choice. As any behaviour in public of said individuals is a matter of display of preference qua conduct in social affairs. As hence anything that has to do with profiling [even if only for the mundane making decisions of what ads to show to certain groups or not; abstracting even from the right (…) to have a human in the loop, seriously], has to do with political preferences.

Where is the field of study, by the way [not so much; rather a both parallel and intertwined track], of metadata and inference being special2 categories of data, not requiring consent but should’ve been outlawed per se ..?

Plus:

[Artful bars, but suppressing; Brittas Museum London]

Appetite for destruction ..?

Not even referring to the Masterpiece. On the contrary, we have here: … Well, what?
Interested as we all are in the subject, since it is defined still so sloppily, we all look for progress, I started. But stopped, when it turned out … risk appetite is defined in hindsight, with a survived disaster being the appetite threshold. Nice. So you’ll know what your appetite is when it hit you and were lucky enough to survive. If you didn’t survive, you now know you passed the threshold. Same [?] with projects: Only if it fails, do you have to write off the investment. The idea of sunk costs may be an enlightenment..?

Etc.

I believe the CRISC curriculum has other, actually somewhat useful, information on this, and on risk tolerance ..?
Your comments, please.

Plus:
[For 20 points, evaluate the risks, e.g., qua privacy, bird strikes, value development; Barça]

Cyber-Allfinanz

Strange, that we see ‘cyber’ (#ditchcyber) Insurance behaving as if it’s not Insurance but banking:
A banker is someone who lends you an umbrella, but wants it back when it starts to rain

Which already has a lot to add; ‘lends against a princely interest sum’, ‘the umbrella will be small, not enough to protect your family that’s the Deluxe edition for a premium’, not ‘starts to rain’ but ‘is predicted to be only slightly possible to have rain in some undisclosed upcoming time period’, ‘wants it back’ means ‘has it reposessed, violently’. Etc.
But that’s not the issue. The issue is that the underwriter of the insurance will not want to pay out. Duh.
Because it’s not if but when you’ll get wet. Despite all reasonable, or more, efforts on your side to protect yourselves from it by not being in the streets when the first drops fall. But then, you can’t stay inside all the time; you’re in business which means going out to play. No matter what sou’wester you don, you’re done.

In other words, no matter how perfect your compliance with, e.g., ISO2700x, you are not safe. Which means you’ve overlooked something, didn’t do e-ve-ry-thing perfectly 100,0% – certainly not when ‘compliance’ means ‘60% or above, of the reasonable efforts’. If the latter is 80% of max, you still end up with having done only less than 50% of what was possible. In the more than 50%, there certainly is something that with hindsight and progressive insight now you’ve been hit you may have done differently.
And the insurers only act on hindsight, qua culpability and cover…

’nuff said; and:
[Differently since positive: Within an unknown Cala hides an unknown Cala; Toronto]

Having fun with voice synth

In particular, having fun the wrong way.
Remember, we wrote about how voice synth improvements, lately, will destroy non-repudiation? There’s another twist. Not only as noted, contra voice authentication for mere authentication (banks, of all, would they really have been in the lead, here, without back-up-double auth?), but in particular now that your voice has also become much more important again [after voice had dwindled in use for any sorts of comms, giving way to socmed typed even when with pixels posts of ephemeral or persistent kinds; who actually calls anyone anymore ..?], we see all sorts of Problems surfacing.

Like, mail order fraud. When hardly anyone still goes on a shopping spree through dozens of stores before buying something in store but rather orders online, of course Alexa / Home/Assistant / Siri / Echo / Cortana are all the rage. For a while; for a short while as people will find out that there was something more to shopping than getting something — but recognising the equilibrium that’ll turn out, may be in favour of on-line business, with physical delivery either at home, or at the mall.
The big ‘breakthrough’ currently being of course some half-way threshold / innovation speed bump overcome, with the home assistant gadgets that were intended to be much more butler first, (even-more-) mall destructor second. But that second … How about some fun and pranking, by catuyrig just some voice snippets from your target, even when just in line behind ’em at Wallmart, and then synthesizing just about any text? When a break-in on the backside of your home assistant (very doable; the intelligence is too complex and voluminous to sit in the front-end device anyway [Is it …!? Haven’t seen anything on this!] so at least there’s some half-way intelligent link at the back) may be feasible per principle but doing a MiM on the comms to some back-end server would be much more easy even, and much easier to obfuscate (certainly qua location, attribution), a ‘re’play of just any message is feasible.

Like, a ‘re’play of ordering substances that would still be suspicious even when for ‘medicinal purposes’. Or only embarassing, like ordering tools from the sort of fun-tools shop you wouldn’t want to see your parents order from. Of course, the joke is at delivery time [be that couriers, DEA/cops, or just non-plain packages] — oh wait we could just have the goods delivered to / picked up at, any address of our liking and have the felons/embarressed only feel that part plus non-repudiability.

This may be a C-rated-movie plot scenario, hence it will happen somewhere, a couple of times at least. Or become an epidemic. And:
[No mall, but a fun place to shop anyway; Gran Vía Madrid]

No surprises here; qua attribution

Is anyone surprised that apparently, “there’s traces of North-Korean involvement” in the WannaCry hiccup ..?
As yesterday’s post (below) already noted; no-one cares about WannaCry1.0 anymore hence ‘hiccup’. Has 2.0 come ’round already?
But how much repudiation by the North-Koreans would reach our general news …? So, how easy it is to blame the NKs for anything that goes wrong ..? Like,

Whereas, … Russia did claim it was also ‘hit’ by WC1.0 [oh the abbrev], but no damage ensued because they were able to stop it at the front door. Right. By lack of actual true snippets from 1600 Penn Ave, we now consider anything that comes out of Russia to be tru-er than what comes from DC, just like that ..? Because that would indeed leave ‘North Korea’ the only reasonably believable/unsurprising culprit.
On the other hand, the embedded tweet indicates Russia actually stole something. Until now, wasn’t it that the exploitable was leaked? Quite different … What is Russia’s involvement now, that those that have info of leakage only, don’t have intel on ..?

[Edited pre-press to add: there… ]

Oh, I’ll just leave it for you to ponder. And weep. And:

[Yes from that ridge, Gettysburg…]

Maverisk / Étoiles du Nord