×

Announcing: Slashdot Deals - Explore geek apps, games, gadgets and more. (what is this?)

Thank you!

We are sorry to see you leave - Beta is different and we value the time you took to try it out. Before you decide to go, please take a look at some value-adds for Beta and learn more about it. Thank you for reading Slashdot, and for making the site better!

Comments

top

Alva Noe: Don't Worry About the Singularity, We Can't Even Copy an Amoeba

justaguy516 Re:writer doesn't get jeopardy, or much of anythin (435 comments)

Nope. A purely empirical observer wouldn't be able to tell you that 'as far as mommy knows, the chocolate is still in the drawer and that is why she is surprised'. The empirical observer would be able to predict that mommy giggles but wouldn't know __why__ she is giggling. The little boy can, because he can model mommy's state of mind. That is the fundamental difference.

yesterday
top

Alva Noe: Don't Worry About the Singularity, We Can't Even Copy an Amoeba

justaguy516 Re:Consciousness versus Intelligence (435 comments)

Computer programs have no idea that what they are tracking is 'time' and how that "time" is different from any other relation (for example, distance) between entities , because they cannot infer causality between events.

Computer programs track a number which changes based on an abstract rule that the programmer programmed into it and based on the absolute value of that number, they do things. You could write a program where this number is replaced by the distance from a fixed point and send the computer on a random walk and the outcome would be very very different, for the same basic rules.

Time is not a sequence of numbers. If I show you a picture of a car down the street and then a picture of the same car in front of you, you will immediately place them in a 'time' order because you infer causality between the two of them......the car is in front of you right now __because__ it was down the street some __time ago__ and it is no longer there because it is here __now__. You think a program scheduler can create a relationship like this spontaneously?

2 days ago
top

Alva Noe: Don't Worry About the Singularity, We Can't Even Copy an Amoeba

justaguy516 Re:writer doesn't get jeopardy, or much of anythin (435 comments)

But all of it doesn't even matter.

The professor of philosophy is actually wrong. We don't understand what time is any better then clocks do. We are complicated, ad infinitum refined "clocks" for mindless, insentient set of genes who aren't even aware of us. Even the notions "aware" and "sentient" are themselves misleading.

No, this is wrong. The professor is right; we do understand time and computers do not. In fact, we are capable of an understanding of time in a way which is impossible to communicate to computers or even any other forms consciousnesses. We understand time because we have an inherent sense of causality built into us, and time is the name we have given to the way we relate causal events. In fact, we are so good with time that we can construct dynamic experiments in our mind, using our own mind as a model and track the state of the modelled mind in time.

Consider the following experiment. A little child watches his mother put a bar of chocolate in the fridge. Little later daddy sneaks in and eats the chocolate. The little boy giggles; why? Because he knows that as far as mommy is concerned the chocolate is still there in the fridge and she will be surprised when she looks for it. How does he know? Because he constructed a simulation in his head of his mommy's mind, fed it a sequence of stimuli and observed its evolution over time. This is a level of manipulation of time (causality) which is completely out of reach of the most powerful computer.

2 days ago
top

As Amazon Grows In Seattle, Pay Equity For Women Declines

justaguy516 Re:I bet Amazon would love to hire more women. (495 comments)

"Amazon" may not care, but "Amazon" doesn't make hiring decisions either. Individuals in Amazon do and they are as subject to societal prejudices as anybody. Not that I know anything about how women are treated in the Seattle tech industry, but this anthropomorphizing of large organizations is pretty silly.

5 days ago
top

Ballmer Says Amazon Isn't a "Real Business"

justaguy516 Re:The US tech industry (283 comments)

Sorry, ARM is doing excellent work as a CPU (design) company. The whole smartphone/tablet business is riding on their coat-tails. So is nVidia.

about a month ago
top

Internet Broadband Through High-altitude Drones

justaguy516 Re:I'm betting on balloons (99 comments)

They do. Otherwise the receivers will have to continuously correct the link timing and Doppler. The link capacity will be degraded.

about a month ago
top

Internet Broadband Through High-altitude Drones

justaguy516 Re:I'm betting on balloons (99 comments)

There was an idea like this floated in the 1990s, called Strato station or something. Balloons at 80km altitude (in the stratosphere) providing coverage. I seem to remember that Loral and Alenia Spaziale were both involved to some extent. It was abandoned because it is too difficult to keep balloons static (even at that altitude) and this would need expensive tracking antennae on the ground (in the 1990s digital beamforming was simply not available for commercial use). Anybody else remember this? It was around the time when the whole world (including Microsoft, remember Teledesic?) was planning to do 'internet in the sky' kind of things.

about a month ago
top

India Successfully Launches Region-Specific Navigation Satellite

justaguy516 Re:GPS (86 comments)

This has something to do with India's recent increased interest in cruise missiles. The Indian Armed forces just recently tested a sub-sonic, nuclear capable cruise missile.

about a month ago
top

Microsoft Lays Off 2,100, Axes Silicon Valley Research

justaguy516 Re: Multiple cuts (109 comments)

Machiavelli: If you have to do bad things, don't do them by dribs and drabs. Do them upfront in one fell swoop.

about 2 months ago
top

Scientists Twist Radio Beams To Send Data At 32 Gigabits Per Second

justaguy516 Re:Spiral filter, and a Tardis (122 comments)

Thanks. This is a little more understandable, once I read the wikipedia entry and saw the phase diagram.

about 2 months ago
top

Scientists Twist Radio Beams To Send Data At 32 Gigabits Per Second

justaguy516 Re:At what signal to noise ratio? (122 comments)

No wireless link operates at an SINR of -40dB. You are mixing up SINR and RSSI. An SINR of 19dB is actually very reasonable; LTE will achieve its top rates only at SINR of 30dB or so.

about 2 months ago
top

Ask Slashdot: Have You Experienced Fear Driven Development?

justaguy516 Re:TDD FDD (232 comments)

Till the point where the 'automated' is considered more important than the 'testing' part and people stop tinkering with the software anymore ; oh, the automated tester will test __everything__.

about 2 months ago
top

Ask Slashdot: Have You Experienced Fear Driven Development?

justaguy516 Re:Fear of changing code.... (232 comments)

That may be, but there are other issues as well. As a tech lead, I frequently hear from a developer, "for feature X, we can either create a brand new state machine, or add to that for existing feature Y; its not too big a deal." Which we finally do also depends on my judgement of this person's capability to make changes to the code-base (it can break a new guy's confidence to be given something too big for him/her to chew, even if they volunteered for it), how much additional testing (regression testing) will be required, whether I need to tell the customer or not (I work in communications software and we can barely test 50% of our feature set in the lab; there are always things happening in the field we don't anticipate).

If we write a new state machine, there may be subtle things that the old state machine handled, which we haven't thought of. On the other hand, if we modify the existing state machine, we may break existing stuff. In either case there is a chance of getting it wrong, but fear causes you to suspend judgement and replace it by paranoia or wishful thinking. Worse, your developers get infected by the same fear and start suspending their judgement.

about 2 months ago
top

Ask Slashdot: Have You Experienced Fear Driven Development?

justaguy516 Fear of changing code.... (232 comments)

[2] is a very common problem, not just because of a badly written code-base, but mostly (IMHO) because of people not having the time to understand a complex piece of code. Ends up in 'nearly' the same code being written in a dozen different places. In my knowledge, it doesn't immediately screw things up, but, over time as the garbage accumulates leads to extremely interesting failure scenarios.

about 2 months ago
top

The MOOC Revolution That Wasn't

justaguy516 Re:Learning is hard (182 comments)

One of the reasons why classrooms work is that the students, knowingly or unknowingly are being constantly evaluated. A good teacher makes a statement (or writes a theorem on the board) and then looks around the class; one look is enough to let her know how many of the students understood, how many didn't and how many (as my father, a professor of EE for 42 years says) haven't even understood that they haven't understood - usually the vast majority. She then restates it, or provides a counter-example, or asks one student to tell her what he understood. This allows the teacher to 'pace' the class. A good teacher also provides breaks within the delivery, to allow the students to sit back and digest what they have heard. This cannot be done fully aposteriori. Each class is different and requires a different pace at different parts of the course. I don't see how this gap is handled through online lectures.

Nevertheless, successful MOOCs have been around for 500 years; they are called textbooks. I am teaching myself Riemannian Geometry using Prof. doCarmo's book as numerous amateurs and auto didacts have done before me. Books, with supporting online forums for specific questions (such as stackexchange) allow one to try and follow at one's own pace, find alternate proofs and alternate explanations of material (for me, it usually takes two good text books).

about 2 months ago
top

When Scientists Give Up

justaguy516 Re:Doesn't surprise me (348 comments)

I was indeed thinking of Giordano Bruno, but he was one of many. The astronomer Cecco d'Ascoli was also burned alive for suggesting that men may live on both sides of a round earth.

about 2 months ago
top

When Scientists Give Up

justaguy516 Re:Doesn't surprise me (348 comments)

You realize of course that there was a time about 500 years back, when scientists were actually burned at the stake for having the wrong theory?

Bernhard Riemann came up with his 'The hypotheses that lie at the fundamentals of geometry' in a lecture which was part of his interview process. He was trying to get a position as a teacher in Heidelberg University where they wouldn't' pay him a salary; just give him a room to hold lectures in and a percentage of the fees that students would volunteer to pay. And this was fairly typical of scientists in the past. Other than a select few, scientists lived in rags, home-taught their kids (since they couldn't afford good schools) and died in penury. Things have been much, much worse for scientists in the past.

I don't want to trivialize the issues that scientists are facing today (my own sister is a scientist and I hear her fights for funding all the time), but please understand that things are way better today than they were in the past.

about 2 months ago
top

Mysterious, Phony Cell Towers Found Throughout US

justaguy516 Re:sensationalism, ahoy (237 comments)

Its unlikely that it is able to "hack into" the base band processor. What it probably does is to pose as a genuine tower and offer to carry the cell phone signals. This then causes the baseband processor to start negotiating with it and bingo, all kinds of stuff is revealed about the phone! That kind of thing would be nearly impossible to keep hidden from the operator; operators routinely do drive by tests using handheld testers and they would know for sure. Given that the operator hasn't done anything about it is fishy.

about 3 months ago

Submissions

top

Aspect Oriented Programming

justaguy516 justaguy516 writes  |  more than 6 years ago

justaguy516 writes "Recently, I was asked to attend a session on aspect oriented programming, which is being rolled out in our organization. I was pretty aghast; we are creating a way to bypass the (imho) the only useful feature of OO. In my mind it is a complete recipe for disaster. We are a software services organization (focussing on the communications domain) and mostly C/C++/embedded in nature, and reliability is a very important aspect; 99.99% availability is a standard requirement. Anybody has any stories/experience to share in this?"
Link to Original Source

Journals

justaguy516 has no journal entries.

Slashdot Login

Need an Account?

Forgot your password?