Wires in your brain?

[bascule]

Which is "trusting user data" without checking?

Memory. Input validation vulnerabilities stem from a failure to remember which data has been properly validated and which has not.

Which is trusting your compiler to not leave too many clues about your encryption cipher implementation?

Not sure what you're referencing here specifically... the RSA timing attack when you fail to use the Chinese remainder theorem?

But I would say unsound cryptography lies in the "math is hard" domain, and you can say similar things about compilers (especially in Lisp and other functional languages where compilation is truly a mathematical transform)

Consider design decisions.

Design flaws are a noticible omission from the list, but I did try to reference them with the complex race condition example.

Consider the notion that even the best coders have a single project limit where they understand all of the pieces at the same time of about 10k lines of code.

And that's a limitation of human consciousness. What happens when you have near-perfect electronically backed memory?

I am not so sure the device would make our math better.

I certainly think it would... the system could either have extensive embedded mathematical capibilities or could attach you to a computer network which could provide them as a service. If you could have the computational abilities of a computer and, say, all of mathworld.com in your head, then suddenly you can compute complex quaternion division rings as easily as you can add 2 + 2.

A sharing of thoughts does not increase the maximum capacity we have to fully understand all the parts of a project in under 10k lines of code.

What is the source of this limitation though? I would say it's a limitation of both short and long term memory...

Now. Consider the number of lines of code in MS Windows XP and estimate how many lines of machine code that would be.

What happens when the visual interface is eliminated and you can jump around inside the code as easily as you jump around your own thoughts and memories?

Consider what would be required for a person to understand the whole project.

You apply an additive effect to "help" the process, but don't apply it to errors. The Mythical Man Month shows us effects of adding people to a project; it does not leaad to linear growth to code productivity. Adding more brains to a project can slow it down. Allowing bad ideas to pass directly? Such effects would be subject to an additive effect as well.

The Mythical Man-Month is more about the limitations of human communication; as the number of individuals working on a given project increases, the communication overhead rises, and so you will always see a point of diminishing returns. Blow away the communication barrier and suddenly you have a conscious network operating from more or less the same input corpus without the traditional barriers of human communication.

I think the best application of transmitting thoughts would be education.

Who needs an education when electronic networks expose your brain directly to unmitigated reality?

[TheCotMan]

Memory. Input validation vulnerabilities stem from a failure to remember which data has been properly validated and which has not.

But a person needs to know this is a risk to begin with. How can a person can't fail to remember something that was never learned?

Not sure what you're referencing here specifically... the RSA timing attack when you fail to use the Chinese remainder theorem?

Several issues WRT to compilers leaving behind information in applications... from static keys, to unstripped products with comments and clues for what was being used to weaknesses in key generation predictability. Again, this is not an issue of trying to remember something if it was never learned.

But I would say unsound cryptography lies in the "math is hard" domain, and you can say similar things about compilers (especially in Lisp and other functional languages where compilation is truly a mathematical transform)

With such a broad meaning for "math" nearly any human task that can be pushed into an algorithm is, "math," and most any human mistakes outside of computers could be argued as being caused because "math is hard."

Overfill a coffee cup? Math is hard. Our system for detecting rate of flow of fluid was flawed, or we did not remember when to stop.
Compilers are not that hard to build-- they are time consuming. When optimization is considered, level of difficulty does increase, but that is not the same thing.

If you want to use such broad meanings (like considering, "building a compiler," a math problem), I might agree with you if I used your definitions.

And that's a limitation of human consciousness. What happens when you have near-perfect electronically backed memory?

That was not part of the original proposal. It is an interesting idea though.

I certainly think it would... the system could either have extensive embedded mathematical capabilities or could attach you to a computer network which could provide them as a service.

This is not the same thing as being able to pass thoughts to a digital network. What you now describe is incorporating digital devices into our consciousness & decision making process, and adding external software/hardware to augment our ability to think. That is not the same thing as using information to make decisions, and isn't the same thing as passing thoughts to a digital network. This is another new item you add. It is also interesting, but not the original proposal.

There is a difference between accessing a digital network with thought and having a digital network as part of our consciousness.

The original statement was one of improving security by allowing abstract thought to talk to a digital network. Though evolution and refinement of your ideas by adding more to them is natural, it makes addressing your original statement more difficult.

What is the source of this limitation though? I would say it's a limitation of both short and long term memory...

I am not sure what the limitation is. If it is memory, then the human mind would still need to remember the existence of an index/key into the external memory storage array. Though this could substantially increase what we could "remember" I am not so sure it could increase what we could consider at one time when looking at a problem.

Consider a metaphor.
A program that uses discrete data in chunks that are 4/5 the size of physical ram, and the OS/application uses the last 1/5. If there are hundreds of chunks of discrete data, we can only evaluate each piece one at a time-- even if we have each chunk stored in virtual memory, paged to disk, or in a file as part of a secondary storage.

Our brains have quite a bit more memory/storage than we are able to use at any one time. Adding external memory does not increase our capacity to keep several memories together while we examine how they work together and actually analyze.

What happens when the visual interface is eliminated and you can jump around inside the code as easily as you jump around your own thoughts and memories?

Then our code may become as reliable as our memory. Consider testimony in court and how different people will have different memories from each other, but a true recording with a camera or electronic recording device does not fail in the same way.

Solution? You might suggest making our memory digital, but digital data suffers from resolution, but that is again not the original proposition.

The Mythical Man-Month is more about the limitations of human communication; as the number of individuals working on a given project increases, the communication overhead rises, and so you will always see a point of diminishing returns. Blow away the communication barrier and suddenly you have a conscious network operating from more or less the same input corpus without the traditional barriers of human communication.

It is not just about communication. It is also about code fixes and time to manage code for a project. It is also about costs of supporting an existing project vs starting a new one. It is about many things beyond "limitations of human communication" or how people are able to put their ideas into a computer.

Who needs an education when electronic networks expose your brain directly to unmitigated reality?

Who needs education when you have access to google? What? ;-)

Seriously, consider bottleneck issues if education is not part of this. People who actually learn things will be "slashdotted" and not be able to much of anything else besides answer queries, and when they go offline? Heh heh. Suddenly, people who have relied upon others to resolve issues assume there are no problems since nobody complained?

What happens when people who were trusted go out and get drunk or are under the influence and then become part of this network of sharing? Or they fall asleep while working on something and start to dream? Couldn't these also cause problems with a project? We also are interrupted with distractions throughout the day. Phone calls, e-mail, peers looking for answers. What happens with shared thought? How much work could you get done if you are constantly being interrupted by different people?
"Hey, does this code look right?"
"Are there problems with this algorithm?"
"What would that co-worker look like if they were naked on my lap?"
"I've got the munchies. Where is the vending machine?"
"Heh-heh. Hrm. Um. Do you have TP for my bunghole?"

You know what? More than education, I see the biggest use of such a system going to a whole new sex industry. People would be porn dealers of thought. You want porn? No need to use your own imagination, use someone else's! You want to have Sex with Paris Hilton? Just download the memories of (ex boyfriend's name.)

There have been movies about stuff like this.
Strange Days

[bascule]

Wow, that's a lengthy ass reply... let me just focus on one part though:

This is not the same thing as being able to pass thoughts to a digital network. What you now describe is incorporating digital devices into our consciousness & decision making process, and adding external software/hardware to augment our ability to think. That is not the same thing as using information to make decisions, and isn't the same thing as passing thoughts to a digital network. This is another new item you add. It is also interesting, but not the original proposal.

It certainly is the original proposal, and if you don't believe it is I think your views of where network services are going are a bit short in scope. When the brain can communicate with digital networks at the level of abstract thought, then any network service would simply become an extension of your own thought process.

The problem is, of course, interface. Once you eliminate the interface barrier between the brain and digital networks, the potential is literally limitless (and I mean that in a transcendental sense). I'm just asking you to consider the full implecations of what such a technology would provide (such as all of humanity merging into a singular conscious entity)

[TheCotMan]

It certainly is the original proposal, and if you don't believe it is I think your views of where network services are going are a bit short in scope.

I did not have the same view as you report here when reading your first post.

When the brain can communicate with digital networks at the level of abstract thought, then any network service would simply become an extension of your own thought process.

I did not want to make that assumption. I could have further assumed other things, like zero latency data transfers, and assume that the speeds to pass thought to the digital network and back to be the same as human thought, but did not want to make these assumptions either.

The problem is, of course, interface. Once you eliminate the interface barrier between the brain and digital networks, the potential is literally limitless (and I mean that in a transcendental sense). I'm just asking you to consider the full implecations of what such a technology would provide (such as all of humanity merging into a singular conscious entity)

It mostly depends on speed (latency) and throughput.

If speed was same as human thought, and throughput was limitless, and there was no resource drain on "clients" or "servers" then you would likley see a geometric increase in the speed at which we are able to solve problems with the mind.
If a person viewed a problem as, "how to bring death to infidels," or, "new proofs in math," or, "how to beat the computer in the game of chess."

Then there are other assumptions. Can new ideas be pushed to people without consent? Is the system only client pull? etc. I can't read your mind... yet. ;-)

However, without methods to enforce limits, it would also create a number of serious problems-- end of the world kinds of problems, risk for loss of diversity problems, mind control kinds of problems.

[LosT]

*Not everyone 'thinks' at the same speed

*Not everyone 'thinks' with the same process/paradigms applied

I still say that connection is not the same a understanding- you may be able to tap into certain facts, pieces of information, etc- but the digestion of such information will always be a function of the person- Just like you can teach someone the tools they need for a complex eplison delta proof, and yet they still may not be able to formulate the proof correctly...

Would learning increase faster if information desired was instantly provided when it was 'requested'? Surely. But you will still reach certain limits imposed by the ability to digest such information.

LosT

[TheCotMan]

*Not everyone 'thinks' at the same speed
*Not everyone 'thinks' with the same process/paradigms applied

True. With shared thoughts, there is risk for elimination of diversity, as people could share the same experiences and experience is a fundamental part of our decision making process.

I still say that connection is not the same a understanding- you may be able to tap into certain facts, pieces of information, etc- but the digestion of such information will always be a function of the person- Just like you can teach someone the tools they need for a complex eplison delta proof, and yet they still may not be able to formulate the proof correctly...

This is the old issue of memorization/recitation vs comprehension. Some of the fastest speed readers may be able to recite facts, and memorize lines, but is this reading for comprehension and retention?

Would learning increase faster if information desired was instantly provided when it was 'requested'? Surely. But you will still reach certain limits imposed by the ability to digest such information.

And we still have an indexing problem (described before.) New ideas are often born of experience and difficulties. Finding new solutions to existing problems is partly a result of genetics and partly a result of experience. If people were "programmed" the same way with the same information, then original invention may suffer. "Think Different" ? It may not be so easy.

[FunkyChicken]

A little OT (and old news)

http://news.bbc.co.uk/2/hi/health/3186850.stm

Something very interesting is a scientific test that was performed in America, Monkeys brains were controlling robot arms over 600 Miles away!!! The monkies had electronic sensors planted in their brain, these sensors would monitor the monkeys brain signals and convert them into what the monkey desired. The monkey would simply control a dot on a computer screen using nothing other then its brain. Scientists linked this up with a robot arm in another lab(600 miles away), the monkey could now perform an action in another state by simply thinking. The reason for these experiments is to enable people who are unfortunate enough to be paralysed to move again.

But how does this relate to this topic? User Interface? If a monkey can control a dot on a screen by thinking, what stops us from controlling a cursor on the screen by simply thinking? This technology opens up thousands of possibilities, in fact, the possibilities are endless, for example, why 'type' a letter when u can simply 'think' it?
We see these technologies in futuristic films, “that'll never happen” you say, change your mind now?

The most important thing to think about in this case is that all of the signals are one way. The monkey can send data by simply thinking, but to recieve data back it must do so by regular senses, i.e. visual.

[mfreeck]

Would learning increase faster if information desired was instantly provided when it was 'requested'? Surely. But you will still reach certain limits imposed by the ability to digest such information.

Here's an interesting thought... Learning something like martial arts... doing anything physical really involves not just book knowledge, but 'muscle knowledge.' Your body learns at a physical level from doing the same motions over and over again. What if you could cut down on that time a bit?

I think it's at a low enough level that it would be either easier or harder. Easier because it's at the brain level, not the mind. OTOH, you'd have to decode those types of 'memories' in order to recreate them and figure out how to transmit them again. I wouldn't be surprised if someone starts working on that for folks recovering from brain injuries.

As far as whether I'd like to be hardwired for communications... No. I don't mind being integrated with external wear, but I draw the line at implants of any sort, with the exception of compensating for a medical deficiency (lack of something I was or should've been born with... eye, arm, well functioning heart).

Regarding pop culture bits being "viruses," I suppose it could be considered a worm. When I think about 'hacking' somebody's mind, I think about brain washing. Replication can be argued both ways... I think it does replicate, but there are few vulnerable hosts. I am speaking more of cults than picking a random individual and forcefully programming them.

[TheCotMan]

...Monkeys brains were controlling robot arms over 600 Miles away!!!

Oh crap! Next thing you know, these monkeys will be web browsing and posting on these forms. That will be a bad day.

[/quote]The reason for these experiments is to enable people who are unfortunate enough to be paralysed to move again.[/quote]
Nice. They have been doing work on eye movement to control operating systems and make "typing" possible by using a special visor or a pair of glasses. Thought thought is a step ahead of using thoughts, both have great potential for those paralized.

But how does this relate to this topic? User Interface? If a monkey can control a dot on a screen by thinking, what stops us from controlling a cursor on the screen by simply thinking? This technology opens up thousands of possibilities, in fact, the possibilities are endless, for example, why 'type' a letter when u can simply 'think' it?
We see these technologies in futuristic films, “that'll never happen” you say, change your mind now?

The idea since before cyberpunk (role playing game) was to have implants for User Interface. Event the idea of the Star Trek transporter may be possible in a distant future. (Consider physics and teleportation in quantum mechanics-- not exactly a transporter, but perhaps research necessary to get a transporter.

The most important thing to think about in this case is that all of the signals are one way. The monkey can send data by simply thinking, but to recieve data back it must do so by regular senses, i.e. visual.

This could also be scary. How many times do we start typing horrible flames on mailing lists only to finally be cool, calm and collected enough to not press "SEND."
Speed of thought? heh. We won't have the same amount of time to cool off before thinking "SEND" :-)

Or, an obviously flawed example:
What if you were thinking, "I wonder what would happen if I formatted my computer's hard disk?"

[enCode]

A little OT (and old news)

http://news.bbc.co.uk/2/hi/health/3186850.stm

Something very interesting is a scientific test that was performed in America, Monkeys brains were controlling robot arms over 600 Miles away!!! The monkies had electronic sensors planted in their brain, these sensors would monitor the monkeys brain signals and convert them into what the monkey desired. The monkey would simply control a dot on a computer screen using nothing other then its brain. Scientists linked this up with a robot arm in another lab(600 miles away), the monkey could now perform an action in another state by simply thinking. The reason for these experiments is to enable people who are unfortunate enough to be paralysed to move again.

But how does this relate to this topic? User Interface? If a monkey can control a dot on a screen by thinking, what stops us from controlling a cursor on the screen by simply thinking? This technology opens up thousands of possibilities, in fact, the possibilities are endless, for example, why 'type' a letter when u can simply 'think' it?
We see these technologies in futuristic films, “that'll never happen” you say, change your mind now?

The most important thing to think about in this case is that all of the signals are one way. The monkey can send data by simply thinking, but to recieve data back it must do so by regular senses, i.e. visual.

You bring up an interesting point. You dont have to make it so your brain IS the computer, rather a medium by which the computer is operated. I think i would do that. Also the medicinal implications are extrodinary.

[bascule]

You bring up an interesting point. You dont have to make it so your brain IS the computer, rather a medium by which the computer is operated.

When you couple high speed bidirectional information exchange directly with the brain with computer networks, the line really begins to blur...

[dYn4mic]

I think If i could control the computer entirely from my mind... (as stated above)
I would grow more fustrated with using the computer... I spend a lot of time wating for the computer to do something, then It doing something for me. This applys to any computer I have ever used (and will use in the next few years you might argue)
An example is using a shitty keyboard or mouse... this interface is normally easy and transparent... but the minute we have a sticky space bar, or a shitty ball mouse... we notice it.. and notice it BIGTIME. It distracts us from working properly and makes our entire computer experance at the time fustrating. (this also applys to mundane details like your car (something that becomes transparent.. A->B) and even security work... (a network intrustion for a person who is trained but has no frequent experiance will make mistakes, be slow and unprepared..etc..etc)

Making my computer input be as fast as I can think it would just be overload (for the most part). The internet is too slow, etc..etc...
I COULD see this in a FPS or gaming... some kind of vision system and sound system I could wear or even "jack in" kinda a deal like the matrix... but then we're just into Virtual Reality.. and processing arch. is noplace near where it needs to be for that.
Or someday when the computer AND the network (world is going to become a giant network... (think of a computer without 'the network' now.. its almost silly in most uses) is as fast as I can think... but then we'll have other things going on.

[TheCotMan]

I spend a lot of time wating for the computer to do something, then It doing something for me.

Some would say that much of the time we wait for the computer has to do with bloated OS [and] applications that are huge-- mostly because of User Interface features that were designed to make use of the system easier and protect users from themselves at a sacrifice of speed.

If the user interface was moved to a hardware interface that meets to requirements [originally] outlined by Bascule, then that would initially lead to an abundance of free cycles due to loss of need for an extensive user interface.

IMO, this abundance of CPU cycles would be short lived. New features would be added to the OS to protect users from themselves.

Wanting to delete certain files, but leave other files behind could be as simple as imagining a regex pattern and having files get selected, or in Bascule's original suggestion, "just thinking about the files" would be enough to have the interface figure it out.

Imagine sharining a computer with a another when you get mad at them and think, "I should delete all their files."

More protections would need to be added as people less techie start to use such interfaces. OS bloat could become even larger than it is now.

[dYn4mic]

That all makes a lot of sense TheCotMan. Maybe this is why my window managers are light, and simplistic themes are choosen (over ultra colorful (etc..))
This is a good quote I've saved from a SecFocus article a while ago...

"I don't believe we'll ever see a reversal of the trend towards bloatware in software. But with security now a major threat to every single software company in the world, the security issues that stem from increasing software complexity must be addressed. The solution to all this complexity cannot be more complexity; we must find a more elegant and simple solution. Security must become ingrained into everything we do."

But, I believe that finding that elegant simple solution is really difficult.
You could argue the number of system calls increasing as evidence that software is moved to bloatware to meet the demand of users.
"UNIX was designed to do one thing, and do it well"

What happends if I use the computer while im intoxicated and my thoughts illogical? What happends when I think about something other than the task at hand? What If i think about deleting a file, but im not sure if I want to or not..?
I don't want a "clippy" in my head, but I also don't want to be deleting or breaking something just because I think about it and the computer moves too fast for me before I change my mind or something.... There's huge implications to something such as this. But TheCotMan is right.. protections would need to get an overhaul and limits applied.

[TheCotMan]

That all makes a lot of sense TheCotMan.

Thanks. :-) Feel free to call me "Cot" since it is fewer keystrokes.

Maybe this is why my window managers are light, and simplistic themes are choosen (over ultra colorful (etc..))

An even more extreme comparison.
MS-DOS vs WIndows XP (Command line vs GUI)
The command line gives experienced users great power.
The GUI gives mediocre users marginal competence.

What happends if I use the computer while im intoxicated and my thoughts illogical? What happends when I think about something other than the task at hand? What If i think about deleting a file, but im not sure if I want to or not..?

Good points.

I don't want a "clippy" in my head...

Gah! That would really SUCK!
"I't looks like you are taking out the garbage. Would you like help with the steps necessary?"

"It looks like you are trying to breathe. Are you sure you want to do that?" (The "[OK]" is greyed out due to a bug. :-o )

"It looks like you are trying to kill that driver on the road in a bit of road rage. Would you like help to do this without getting caught?"

"It looks like you are posting on usenet, would you like help with flaming and trolling?"

Gah! The humanity!