Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - Cramulus

Pages: 1 [2] 3 4 5 ... 769
16
okay, ramble time


https://techcrunch.com/2016/11/22/googles-ai-translation-tool-seems-to-have-invented-its-own-secret-internal-language/

As DNN's and the like become more complex and learning increasingly unattended, more and more their operations become unfathomable to meatware. Potential for lulz are off the charts.

Q) "Skynet - why did you nuke Sweden?"
A) "#1#B:)"w8;,1"lk"



So, Machine Learning... first, the sci-fi "are machines going to outsmart us?" angle -----

My (mis)understanding is that these things work by observing the existing relationships within a corpus. When asked to make a decision, it's just selecting the "most probable response" based on the existing relationships.

To me, that implies a ceiling of what's capable using this method. Neural networks trying to model human intelligence can become, at best, as smart as a human. If there is some advanced form of reasoning that we don't use, it won't appear in a neural network (at least, not one that's studying humans).


Next, let me talk about this clickbaity TechCrunch article... Lemme see if I understand.

Google wants to translate from, say, Japanese to Kivunjo. It has not been programmed with the explicit relationship between Kivunjo and Japanese words. But it can figure out through context that the word "dog" in japanese is XXX and the word "dog" in Kivunjo is YYY and then translate XXX into YYY.

Quote from: the article
...does that mean that the computer has formed a concept of shared meaning for those words, meaning at a deeper level than simply that one word or phrase is the equivalent of another?

Kinda.. in the same sense that google's autocomplete 'understands' what you're looking for. It's extrapolating based on context clues. Whether you merit that as 'thinking' or 'just following a smart algorithm' is up to you.

Quote
In other words, has the computer developed its own internal language to represent the concepts it uses to translate between other languages?

I don't know that I'd credit the computer with that kind of agency. I'd phrase it like - the programmers of the google translation tool developed a really interesting semantic engine. It uses metadata like context and grammar to guess the translation of any given word, without being given an explicit dictionary.



let me hack at this with a different axe



-"I love to pet my Cujo. He has four legs and wags his tail when he's happy. He barks when he's angry."
-"I love to pet my Breenbal. He has four legs and wags his tail when he's happy. He barks when he's angry."

After being fed these sentences, the google translation bot will recognize that the word Cujo and Breenbal are used in the same context. When you see one of those words, you're also likely to see words like "pet", "four legs", "wags tail", "bark"... The computer may flag this as a likely synonym for "dog".

When the computer is generating a sentence, it might use Cujo or Breenbal interchangeably. This is because they have similar 'semantic webs'. We didn't need to teach it that Cujo = Dog and that Cujo = Breenbal because it gets its meaning from context. Both terms commonly appear with words like "pet", "bark", etc.




Is it "emergent behavior"? Mmm I wouldn't call it that.

Are machines getting smarter? Eh, I think programmers are getting smarter, and machines are using more complex techniques, but it's not like this thing actually understands the 'meaning' of these words.

Is this an extremely clever approach to translation? Very much so.

17
The Secrets Forum / Re: Space dogs
« on: November 30, 2016, 05:52:39 pm »
Hi everybody!   :wave:   Interesting thread...

If we encounter a planet with no life on it, and no obvious 'potential for life', I think it's "fair game" for mining and exploitation.

I think that planets which may one day have life need to be approached cautiously. But then the question becomes - When does life [on a planet] begin? At something like amino acids?


In my daydream hypothetical contact with alien life, I imagine the definition of 'life' will come into question. Like maybe life on other planets don't consist of distinct organisms, but something like distributed systems with lifelike properties.

There's a really good Kim Stanley Robinson book where space explorers have to deal with sickness caused by these alien 'prions'. It's basically just a protein molecule that unfolds in a dangerous way when exposed to some of the chemicals in our bodies. Like a virus, we don't consider a prion to be living... but when you're on an alien planet, how can you be sure?

I conjecture that we can (and should) come up with an agreement about ethical exploration.. but I think that parts of the plan will be confusing to apply. Whether a planet has 'life' or not may be fuzzy, hard to determine until we're already knee deep.

18
The Secrets Forum / Re: McConnelling.
« on: June 09, 2016, 03:26:08 pm »
ohhhh yeahhhh

*makes eye contact with the camera like Ferris Bueller*

https://www.youtube.com/watch?v=AxzmV1CcVv8


19
The Secrets Forum / Re: High Altitude Hell, part 2
« on: June 09, 2016, 03:13:41 pm »
That was awesome

20
Techmology and Scientism / Re: Plant Watering Reminder Gadget
« on: June 09, 2016, 03:09:55 pm »
I built this thing at my desk at work. No less than four people asked me if what I was building was "dangerous".

-Two of them were asking tongue-in-cheek
-The other two seemed weirdly hopeful that it was

21
Techmology and Scientism / Re: Plant Watering Reminder Gadget
« on: June 07, 2016, 08:12:09 pm »
Arduino Uno + 3 LEDs + 3 buttons + 6 resistors

22
There was, and still is, option 3 - Never involve yourself with it to begin with.

well you are posting on a forum, so..  :p

But yeah, this is the minimum amount of privacy you can trade for community.


I like your comparison between engaging in social media and committing to watch 150 hours of a TV show. Do you need it? Is teaching you good messages? Is it worth the sacrifice?

Me, I got a FB account years ago when it was the best way to communicate with a girl I was interested in. And then I found out I had been missing parties my friends were throwing, because they were only sending out invites via FB. So it was immediately worth it to me. But there, I'm really just talking about its utility in facilitating local communication.

I do find I have a number of friends who I see like 1x per year, and reading their damn status updates every day helps me feel like I "know" them. When I do see them in person, our relationship feels richer, we have more to talk about, I understand them better. So that's a benefit, for me at least.

But it's really hard to measure the drawbacks - the reenforcement of norms, the bubble effect, the groupthink... a lot of the drawbacks are intangible, or are just levels of risk.

23
Techmology and Scientism / Plant Watering Reminder Gadget
« on: June 07, 2016, 04:39:01 pm »


The plants in my apartment kept dying, mainly because my girlfriend and I kept assuming the other one had watered them. So I invented a little gadget that tells us when the plants need to be watered. Itís super simple.

The gadget has 3 buttons and 3 LEDs. Each pair is linked to one of the plants in my apartment.

Itís only a timer, it doesnít detect soil moisture or anything crazy like that. When a light turns on, it signifies that one of the plants needs to be watered. When somebody waters it, weíll press the button, and itíll reset the counter.

This fun little gadget only cost about $20 to make. It was a lot of fun to build!

24
Literate Chaotic / Re: And the Mountain Never Got Closer
« on: June 07, 2016, 04:13:04 pm »
:mittens:

25
Think for Yourself, Schmuck! / Laws, Social Media, and the Panopticon
« on: June 07, 2016, 04:04:33 pm »
One of the interesting things Junkenstein touched on in the Laws thread is the the intersection of law enforcement and surveillance.

The current incarnation of the justice system is based on some ideas that were kicking around the post-monarchial world and went through the wash a few times until they came out clean enough to sell. Namely the Panopticon.

The idea is that bad people will keep doing bad things when the authority isn't looking- so you should put prisoners in a place where they can be observed. Maybe it's that their prison cell faces a 1-way mirror and there might be a guard on the other side who gives them a demerit if they don't act right. Or maybe it's because they're picking up trash by the side of the road and they are constantly feeling the judgment of the public.

But the effect of observation is that it forces the person - the target of the discipline - to internalize a set of rules. They have to behave in a certain way to avoid reprimand, and eventually they internalize these rules. That's how you affect somebody's spirit, you make them build the laws inside of them, going through the motions with their body. Even if they're faking it, still thinking criminal thoughts, eventually this will be eroded by routine and discipline.

And that's how we built schools, prisons, the military...

And now we're in the social media era, everything can be uploaded, commented on, upvoted and downvoted.

There was a twitter account that this guy started on a subway in NYC, it was just supposed to document people who were taking up too much space, or doing asshole things that aren't appropriate for the subway. He wanted to shame people. And it turned into this really acrid account where he was just picking on people for the way they were dressed, etc. And eventually, the backlash came to him - somebody figured out who he was, and got his pictures up on Twitter. Then, anybody who saw him in public would take a pic and tweet it at that account. He used his phone camera as a weapon, and discovered it was a double-edged sword.

What rules are being enforced by the social media panopticon?

The terms of the transaction are: you trade some your privacy for access and community.

Years ago, I thought the way to deal with this was twofold:

1. Maintain the ability to disappear completely. At any time you should be able to kill your account and escape from whatever storm is chasing you. To do this successfully, you should avoid using your real name on the net. You should keep personal details obscured, and don't leave any channels open that you can't later close.

2. Chaff. Cover your social media with false cues and information. This way, anybody datamining you will get confused and the value of any given data will come under question.


Are those still good strategies?

My real name is on FB, it's easy to track me down... it's a bit harder to figure out my real name from here, so maybe the ability to disappear completely is more contingent on (a) how well you've compartmentalized, and (b) how motivated people are to connect the dots.

If I publish Chaff, it kind of gets in the way of the point of using social media to begin with.

and as I think about the steps you have to take to escape an angry wasps nest, I think to myself... what rules do I have to follow to avoid provoking the wasps to begin with? what rules has this panopticon made me internalize?  What rough edges have I actually sanded off, not just concealed? Has the presence of this social power refined me like a crucible, or restricted me like a warden supervising a chain gang?



28
Help me locate a thread... I recall there being a great thread around here where somebody was talking about That Good Anarchist Utopia and people smacked some sense into him. Actually there's probably a dozen threads like that.

Where did the "I'm gonna take your sandwich" meme come from?

 I need to smack some sense into a deluded anarchist and I think that thread might have been a right and proper Chair Leg of Truth.

29
I like LMNO's framing there.

I feel like a lot of the objection to common core is that the adults in this metaphor always used food processors and can't fathom their kids being taught anything different.

30
The Secrets Forum / Re: PICS VIII: 10% LARGER THAN PICS VII
« on: May 11, 2016, 04:47:07 pm »

Pages: 1 [2] 3 4 5 ... 769