Slashdot Mirror

That statement should apply to both self-driving and human cars

No, because they're not the same at all. Self-driving cars are in many respects attractive nuisances. If you don't think they will be tempting targets for malicious interference (e.g., spray-paint cameras or laser pointers aimed at cameras), then you haven't thought about it enough.

I understand that Waymo and others have done a lot of testing and undoubtedly have their own sets of standards they think are sufficient to ensure a reasonable level of safety, but I don't have access to those standards or the test data, so I have no way of assessing the results. What I don't see are the automotive equivalent of FAA regulations (FARs) or airworthiness standards (e.g., MIL-STD 516).

"The US Army is looking to design things based on thousands of soldiers playing video games and lego-blocking up designs. When they figure out what works best in the game, they 3d print the thing. http://www.dau.mil/publication...

I work for the federal Government, and I would just like to say that this is false. We do not go with the lowest bidder on many occasions. In fact, here are some actual documents on the subject:
description of best value process - https://acc.dau.mil/CommunityB...
acquisition regulation mandating best value - https://acc.dau.mil/CommunityB...
Federal Acquisition Regulations indicating best value - http://www.acquisition.gov/far...

I work for the federal Government, and I would just like to say that this is false. We do not go with the lowest bidder on many occasions. In fact, here are some actual documents on the subject:
description of best value process - https://acc.dau.mil/CommunityB...
acquisition regulation mandating best value - https://acc.dau.mil/CommunityB...
Federal Acquisition Regulations indicating best value - http://www.acquisition.gov/far...

The DoD already has access through contract to that software. The problem isn't access / purchase of the software, the limitation is the security paperwork needed to USE any of that software! ( https://acc.dau.mil/CommunityBrowser.aspx?id=22645 ) The security paperwork that is required can be long, very long, sometimes HUNDREDS of pages long, and take *YEARS* to get reviewed and approved! The DoD just keeps ADDING bureaucratic layers to this process every year as well!

Why does this make it seem that the US military is becoming more and more like the Vogon armed forces.

The DoD already has access through contract to that software. The problem isn't access / purchase of the software, the limitation is the security paperwork needed to USE any of that software! ( https://acc.dau.mil/CommunityBrowser.aspx?id=22645 ) The security paperwork that is required can be long, very long, sometimes HUNDREDS of pages long, and take *YEARS* to get reviewed and approved! The DoD just keeps ADDING bureaucratic layers to this process every year as well! There's a point where the security paperwork just causes more harm than good. By the time you get the software solution engineered and approved, its already most of the way to being completely obsolete! You want to fix the software in the DoD? Fix the process that governs it! Streamline it, cut out the what has by now become multiple layers of unneeded CRAP that's only there because a spot failed at some point, and the solution they came up with simply involved just adding more layers to an already unruly behemoth!

TL;DR - Good luck M$! By the time you get Windows 8 approved, it'll be 4-8 years later.

Even if the specific example may have been blown way out of proportion, I actually see a lot of the plainly-worded outrage as a complete misunderstanding of how the whole DoD acquisition process actually works.

To this end, I once wrote up a notional piece on The Mythical $800 Hammer. :-)

Admittedly I have no experience with city contracts and living in a different country I don't know much about american government's procurement process but it seems to me that all their deals will be put under a microscope from now on. I wouldn't be surprised if they lost their government contracts anyway, regardless of the fine. Fact that an actual huge multinational corporation has actually been found guilty is what gives me hope.

They technically have not been found guilty of anything. had they been tried and found guilty, then they could have been barred from receiving new contracts but a settlement doesn't do that. At best some agencies may look at this and have second thoughts; but even so it may be tough, under the FAR (Fed Acquisition Regs)*, to not award them a contract on which they are the qualified low bidder simply because of the NYC settlement.

More than likely it will have little, if any, impact on SAIC's ability to win contracts. Oteh regencies will probably chalk it up to the typical implementation spat when a s/w project has problems.

* If you really want a sense of how Byzantine the FARs are; drop by the Defense Acquisition Portal https://dap.dau.mil/Pages/Default.aspx

Most contractors merely charge the govt $150 - 250 an hour for the same people the govt uses already, while at the same time, carrying little risk. Compare this to a GS-14 at less than a $100 an hour, inclusive of all costs.

You'll see a move to contract types* ** like cost plus, or cost plus fixed fee, where the government pays out the nose for cost overruns on the part of contractors. Fixed price contracts will only be made with massively inflated rates in order to protect contracting firms from risk.

This leads to massive poaching of govt personnel to the private sector, and vastly inflated rates to the govt.

The privatization of the US government is an abject failure. A-76*** is an abomination, because it does not consider the long term efficiency by private vs public sector.

* http://www.dtc.dla.mil/dsbusiness/Info/contracts1.htm
** http://www.dau.mil/pubs/misc/toolkit.asp
*** http://www.whitehouse.gov/omb/Circulars_a076_a76_incl_tech_correction/

Correct link: http://www.dau.mil/pubs/IDA/chart%20front.pdf ... and I have to say: wow. This is why military projects start at $billions and go up from there.

And there's a reason all that stuff is in there. About 90% of the stuff on there is error checking, accountability and oversight. A small part of that is what I do for a living. Yes, it's an unwieldy chart (I have a paper copy, it's stupidly large), but some very smart people have developed that over many years. If you have any specific complaints or questions on the chart, there's a good chance I can answer them.

This chart is a classroom aid for Defense Acquisition University students. It provides a notional illustration of the interfaces among the three major decision support systems used to develop, produce, and field a system for national defense. Defense acquisition is a complex process with many more activities than shown here, and many concurrent activities that cannot be properly displayed on a two-dimensional chart. Supporting information is on the back of this chart. For more information see the Acquisition, Technology & Logistics Knowledge Sharing System (http://akss.dau.mil).

This chart is a classroom aid for Defense Acquisition University students. It provides a notional illustration of the interfaces among the three major decision support systems used to develop, produce, and field a system for national defense. Defense acquisition is a complex process with many more activities than shown here, and many concurrent activities that cannot be properly displayed on a two-dimensional chart. Supporting information is on the back of this chart. For more information see the Acquisition, Technology & Logistics Knowledge Sharing System (http://akss.dau.mil).

To better understand it, you should read the explanations in the backside of the chart. Awesome!

Also, don't forget that anything major project is managed according to this chart. :-)

Now the fun part... Try and find the boxes in the diagram where something functional actually gets built!

Correct link: http://www.dau.mil/pubs/IDA/chart%20front.pdf ... and I have to say: wow.

This is why military projects start at $billions and go up from there.

Also, don't forget that anything major project is managed according to this chart. :-)

Now the fun part... Try and find the boxes in the diagram where something functional actually gets built!

Correct link: http://www.dau.mil/pubs/IDA/chart%20front.pdf ... and I have to say: wow.

This is why military projects start at $billions and go up from there.

Also, don't forget that anything major project is managed according to this chart. :-)

Now the fun part... Try and find the boxes in the diagram where something functional actually gets built!

This is what most people here simply don't know about DoD/Gov't employment. The people who work for the gov't in that world aren't doing interesting technical work. They're managing projects at a high level, sifting through requirements, sitting in meetings, and setting up contracts.

Oh, and they've also taken lots of excruciatingly boring courses on understanding this process. (ok, DARPA gets an exemption from that, but everyone else doesn't)

Whenever you hear about a cool new DARPA/DoD project, its not the DARPA/DoD folks who are actually doing the cool work. Its non-gov't people working for some company the gov't has a contract with that actually have all the fun.

He is a myth. Again, look up the stats. Very few individuals have the money to first of all, hire a patent bureau to draft a patent application. Unfortunately, the USPTO accepts almost any crap applications these days so they won't be properly screened for prior art. Thus, what this inventor gets is not a patent but a time bomb. As soon as he tries to use his patent in any way, he will be slapped with so many infringing law suits his head will spin fast enough to create its own magnetic field, giving him cancer.

You eliminate any incentive for independent inventors to invest capital in their projects

Several large VCs claim to value patents very low compared to many other factors when deciding to invest in start-ups. Factors they value much higher include time-to-market and their ability to execute and ship product before the competition. Patents tend to slow that process down.

For companies who are working in a patent riddled space, I definitely do a mental calculation of the added risk of litigation and subtract that value from the valuation of the company or decide to not invest at all.
Joe Ito, VC

No, this is what PATENTS do, today. Since it's become impossible for lone inventors to create proper patent applications, they need the resources of very large patent bureaus to at least make the attempt to create non-infringing new patents. Again, this is a crap shoot, at best.

We have this really smart guy in Sweden named Håkan Lans. Maybe you heard of him? Look him up. He got a patent for a small invention that turned out to be really useful for color graphics. He got hold of a reputable patent bureau and started raking in the licensing fees until one day some large corps basically told him to fuck off and die. He tried to fight back, but the sneaky bastards bought off his patent lawyers so they made a rookie mistake in filing (yeah right), he lost on the technicality and he now owes the law firm somewhere around 10 million dollars (estimates vary upwards from that, I've seen speculation about a hundred mil). They say they'll back off if he gives them another one of his patents. He hasn't been able to work and invent since 1999 because he's been tied up in this crap that you claim would PROTECT him? Excuse me while I go out the back and laugh myself silly. The patent system is rigged against the small inventors. The big patent bureaus and corporations are selling pre-scratched lottery tickets through the USPTO. It's all a big myth.

Here's one where Henry Ford claims he did no such thing as invent the assembly line and the whole "lone inventor" notion is a myth:
Bridging Small Worlds to Accelerate Innovation (PDF)

I recently went to a talk by a guy (Dave Brown of DAU and GMU ) who got his PhD recently, and used Legos in his dissertation experiments. He showed that by "learning" a Bayesian network from actual performance data of a system you could create a model that would predict the performance of the system much more accurately than the textbook formulae it was theoretically supposed to follow.

To show this he studied battery decay patterns by running lego models around and measuring the speed they went as they ran out of juice. He also uses lego models for prototyping in the classes he teaches at Defense Acquisition University.

In short, this guy gets to play with legos at his paying job, and for his PhD project. The bastard. I'm so envious. I gotta figure out how to work that into my job.