Locked Out Of The Sixth Amendment By Proprietary Forensic Software
from the unconfrontable-witnesses dept
There are all sorts of proprietary code out there, unavailable for inspection and perfectly capable of sending people to prison, as Slate's Rebecca Wexler reports.
Defendant Martell Chubbs currently faces murder charges for a 1977 cold case in which the only evidence against him is a DNA match by a proprietary computer program. Chubbs, who ran a small home-repair business at the time of his arrest, asked to inspect the software’s source code in order to challenge the accuracy of its results. Chubbs sought to determine whether the code properly implements established scientific procedures for DNA matching and if it operates the way its manufacturer claims. But the manufacturer argued that the defense attorney might steal or duplicate the code and cause the company to lose money. The court denied Chubbs’ request, leaving him free to examine the state’s expert witness but not the tool that the witness relied on.That's a starkly mercenary stance to take. The "trade secret privilege" invoked here basically states that the company's potential loss of income outweighs a person's potential loss of freedom. It also asks for a level of trust it hasn't earned: that the software is as close to infallible as it needs to be. Cross-examination is next to useless when the software itself can't be examined.
Worse, this closed-off software operates in a field where nearly every previous form of "indisputable" evidence has proven to be severely flawed.
Studies have disputed the scientific validity of pattern matching in bite marks, arson, hair and fiber, shaken baby syndrome diagnoses, ballistics, dog-scent lineups, blood spatter evidence, and fingerprint matching. Massachusetts is struggling to handle the fallout from a crime laboratory technician’s forgery of results that tainted evidence in tens of thousands of criminal cases. And the Innocence Project reports that bad forensic science contributed to the wrongful convictions of 47 percent of exonerees.Everything tied to securing convictions seems to suffer from pervasive flaws compounded by confirmation bias. For four decades, the DOJ presented hair analysis as an unique identifier on par with fingerprints or DNA when it wasn't. A 2014 Inspector General's report found the FBI still hadn't gotten around to correcting forensic lab issues it had pointed out nearly 20 years earlier. This contributed to two decades of "experts" providing testimony that greatly overstated the results of hair analysis. All of this happened in the FBI's closed system, a place outsiders aren't allowed to examine firsthand.
That's the IRL version. The software version is just as suspect. Computers aren't infallible and the people running them definitely aren't. If the software cannot be inspected, the statements of expert witnesses should be considered highly dubious. After all, most expert witnesses representing the government have a vested interest in portraying forensic evidence as bulletproof. Without access to forensic software code, no one will ever be able to prove them wrong.
If a piece of software has the ability to deprive a member of the public of their freedom, its code should be open for inspection by the defense. "Trade secrets" should not take precedence over the public's right to defend themselves in court. Even in the highly unlikely event that Chubb's defense team would have copied the code and destroyed the company's future profits, it would still have the ability to seek redress through the court system. After all, that's the line the government uses when it argues for expanded "good faith exceptions" or warrantless searches and seizures: "Hey, if we screw up, you can always sue."
The judicial system is a remedy for wrongs, both criminal and civil. What it shouldn't be is a protective haven where ridiculous assertions like those made here are used to prevent an accused person from learning more about the evidence being used to convict them.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: accuser, code, devices, forensics, lawsuits, sixth amendment, software
Reader Comments
Subscribe: RSS
View by: Time | Thread
[ link to this | view in chronology ]
[ link to this | view in chronology ]
Intellectual Property Trumps Constitutional Rights?
[ link to this | view in chronology ]
Re: Intellectual Property Trumps Constitutional Rights?
(Remember all those debates about why software isn't patentable because it's just a thought process, which explicitly isn't patentable? Same thing applies here.)
If the witness can't explain his thought process (as enacted through instructions to a general purpose computer) to the jury that is grounds for questioning his credibility. Same as if he declared the accused to be a witch but couldn't explain why.
[ link to this | view in chronology ]
Re: Re: Intellectual Property Trumps Constitutional Rights?
V: There are ways of telling whether she is a witch.
P1: Are there? Well then tell us! (tell us)
V: Tell me... what do you do with witches?
P3: Burn'em! Burn them up! (burn burn burn)
V: What do you burn apart from witches?
P1: More witches! (P2 nudge P1)
(pause)
P3: Wood!
V: So, why do witches burn?
(long pause)
P2: Cuz they're made of... wood?
V: Gooood.
(crowd congratulates P2)
V: So, how do we tell if she is made of wood?
P1: Build a bridge out of her!
V: Ahh, but can you not also make bridges out of stone?
P1: Oh yeah...
V: Does wood sink in water?
P1: No
P3: No. It floats!
P1: Let's throw her into the bog! (yeah yeah ya!)
V: What also floats in water?
P1: Bread
P3: Apples
P2: Very small rocks
(V looks annoyed)
P1: Cider
P3: Grape gravy
P1: Cherries
P3: Mud
King: A Duck!
(all look and stare at king)
V: Exactly! So, logically...
P1(thinking): If she ways the same as a duck... she's made of wood!
V: And therefore,
(pause & think)
P3: A witch! (P1: a witch)(P2: a witch)(all: a witch!)
[ link to this | view in chronology ]
Trade secret?
So, it's your choice - go to prison for a murder or for a an attempt to decode a piece of software. And you never know ahead which sentence is longer...
[ link to this | view in chronology ]
The flag of the People's Republic of China goes up in front of our home tomorrow. No doubt later in the day the Stasi will kick in our door and we will never be seen or heard from again. Just drop all of your ordinance on my position, live free or die.
[ link to this | view in chronology ]
Re:
...Wow.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re: Re:
Jack
[ link to this | view in chronology ]
That totally makes sense.
[ link to this | view in chronology ]
there are no surprises
[ link to this | view in chronology ]
Re: there are no surprises
[ link to this | view in chronology ]
Re: Re: there are no surprises
The underfunding of public defenders doesn't help. But, you tell a jury this guy in a suit can prove it through means they don't really understand, and then they'll think back to all those nights they fell asleep watching CSI, and think, "well I trust these guys, they've got magic science to back them up."
[ link to this | view in chronology ]
Re: Re: Re: there are no surprises
Asking for a conviction based solely upon DNA is a stretch. Is there no circumstantial evidence to support the charges? For example, was the dude even in town at the time? Did he use a credit card? The lack of supporting evidence is ridiculous and s bit suspicious.
[ link to this | view in chronology ]
Re: Re: Re: Re: there are no surprises
[ link to this | view in chronology ]
Re: Re: Re: Re: there are no surprises
That works about as well as instructing a jury to ignore something a prosecutor/defense attorney said that they weren't supposed to.
[ link to this | view in chronology ]
Simple Math
https://www.schneier.com/blog/archives/2009/05/software_proble.html
[ link to this | view in chronology ]
Re: Simple Math
[ link to this | view in chronology ]
Re: Re: Simple Math
50% of people are below median intelligence, but we can't assume intelligence has a normal distribution. The way IQ is calculated assumes a normal distribution, but that is just a trick of the math.
[ link to this | view in chronology ]
Re: Re: Re: Simple Math
[ link to this | view in chronology ]
Re: Re: Re: Re: Simple Math
Those that understand binary, and those that don't.
[ link to this | view in chronology ]
Re: Re: Re: Simple Math
They are a good predictor of how well you'll do in college, and can identify people with certain cognitive problems, but that's about it.
[ link to this | view in chronology ]
Re: Simple Math
in other words "let us look at the code or we will assume it is just like this pile of crap"
[ link to this | view in chronology ]
Just a thought.
[ link to this | view in chronology ]
Re: dna testing
[ link to this | view in chronology ]
Aargh!
[ link to this | view in chronology ]
[ link to this | view in chronology ]
Response to: Anonymous Coward on Oct 8th, 2015 @ 7:12pm
[ link to this | view in chronology ]
Re: Response to: Anonymous Coward on Oct 8th, 2015 @ 7:12pm
[ link to this | view in chronology ]
Otherwise, it's only fair to allow the defense to use their own black box software. Potentially exonerating the defendant of any wrongdoing.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
The shareholder salute
That would paradoxically result in rendering the same court invalid, which may be the goal of the corporate grab for power. Its the investor-state dispute settlement model, just transmogrifed into investor-citizen dispute settlement.
[ link to this | view in chronology ]
believing is not science
the it is NOT science
like
"social sciences"
"women sciences"
"political sciences"
or in Peter Thiel words:
“Whenever you can’t have a debate, I often think that’s evidence that there’s a problem,” Thiel said on The Glenn Beck Program. “When people use the word ‘science,’ it’s often a tell, like in poker, that you’re bluffing. It’s like we have ‘social science’ and we have ‘political science,’ [but] we don’t call it ‘physical science’ or ‘chemical science.’ We just call them physics and chemistry because we know they’re right.”
[ link to this | view in chronology ]
Re: believing is not science
[ link to this | view in chronology ]
xkcd
https://xkcd.com/683/
[ link to this | view in chronology ]
What's wrong with a pen and a piece of paper?
[ link to this | view in chronology ]
Re: What's wrong with a pen and a piece of paper?
[ link to this | view in chronology ]
Re: Re: What's wrong with a pen and a piece of paper?
[ link to this | view in chronology ]
Re: What's wrong with a pen and a piece of paper?
[ link to this | view in chronology ]
The legal system REALLY hates admitting it made a mistake, to the point that they will execute someone proven to be innocent.
The system punishes those who manage to prove they were railroaded. See also: Central Park Five, West Memphis Three, McMartin Preschool, and more.
Imagine the pressure brought to bear that if you allow this person to examine the evidence you might undo convictions of all sorts of "bad people" who were found guilty... but only because of flawed evidence.
We treat the legal system like a game, we reward for wins only. This is why we see prosecutions declined, the case isn't a slam dunk and a loss would hurt the score. We can't indict the officer, other officers might not play nice with us any more. Oh this idiot, we can just pile up the charges and tell him hes facing life in prison but if he takes our deal... and boom conviction and another gold star on the board.
This is no longer justice for all, it is a game and one side has all of the power and we have seen they think nothing of condemning an innocent to the ultimate penalties to keep their records secure. This broken system is what we helped build, it will be hard but we need to tear it down and go back to what it was supposed to be.
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re: Re:
Man shoots man who did something bad himself.
There is nothing wrong with offering slightly better terms to hasten this case through.
Justice can be served.
However if there is a reward to just stack & stack & stack charges and turn what would have been a 5 yr sentence (guesstimate) into threats of a 50 yr sentence to get him to accept 10 so the prosecutor gets a better gold star its broken.
Do you think the public would be perfectly okay with a system where they had to openly admit we found as many charges as we could to get him to accept these terms?
The justice system is broken, the fact that we need to have deals to keep the system from collapsing raises questions about have we gone to fing far. Are we pursuing people in a silly game of we passed these stupid laws to make stupid people feel safer, yes we are. We have panic driven laws - see also powder cocaine vs rock cocaine, where there shouldn't be a damn difference but because of nonexistant crack baby hysteria sweeping the nation there is.
Once they have served their time, we still have these awesome systems designed to keep punishing them, so we can feel safer. We have petty tyrants at every level inflicting more misery because their targets can't fight back. Look at the Judge who threw a domestic violence survivor in jail because she got away from him and didn't want to relive it one more time. I think it was 6 prior DV charges weren't enough for him to be found guilty and the Judge got on her high horse of respect my authoritah! and into jail the victim goes.
When the legal system denies someone the ability to prove or disprove a tool works, it is failed. You can cross examine the expert but even he can't answer how the damn thing works... but the CSI effect means people believe the machines are perfect.
[ link to this | view in chronology ]
[ link to this | view in chronology ]
[ link to this | view in chronology ]
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re: Re:
And humans can make mistakes, people know and accept this... the machine is always viewed as perfect and showing it had any flaw shatters trust in its findings.
[ link to this | view in chronology ]
In cases like this
[ link to this | view in chronology ]
Re: In cases like this
Or, just basically flawed from the get go. Intel was selling processors for years which couldn't do floating point arithmetic correctly. Microsoft's calculator app, ditto.
This sounds a lot like the abysmally flawed voting machine debacle. However with your life potentially on the line, the response in this case is nonsensical. I wonder if in ten years we'll hear of a class action lawsuit by the wrongfully convicted (or their heirs) suing this company into oblivion.
[ link to this | view in chronology ]
I also agree that the only downside of the release of the source is that the program will prove to be seriously flawed and thus cost the company customers and revenue. A program used to convict and possibly kill a man must be open to examination by the defense experts.
[ link to this | view in chronology ]
Re:
Which means that if the death penalty is given the case is automatically appealed. Then the higher court(s) can look at the issue and possibly judge a procedural error.
[ link to this | view in chronology ]
Re:
In fact, that program if used, ought to be developed and controlled and owned by the gov't for use by the justice system. NIST, maybe? No defence lawyer should have to defend against, "'Cause we said so, so there." That's the opposite of blind justice. It's divine right of kings.
[ link to this | view in chronology ]
Exhibit 1 VW
[ link to this | view in chronology ]
And that right there is the real reason they won't release the code. It's not because of trade secrets, it's because they (and the government) are afraid of losing their edge.
[ link to this | view in chronology ]
If you have done nothing wrong you have a lot to fear.
[ link to this | view in chronology ]
Vangelder, Camblin, TrueAllele
People v. Vangelder
In June 2015, Mass. ordered a hearing on the reliability of a newer blood alcohol testing device.
Commonwealth v. Kirk P. Camblin
Defendants have argued that the source code used by the device is flawed.
This after Judge Maureen Clancy denied discovery of TrueAllele source code for DNA analysis in Ohio in 2014.
...
[ link to this | view in chronology ]
If (defendant)=not_white {goto guilty}
if (defendant)=poor {goto guilty}
else {goto innocent}
end if
function Guilty()
{call prison-slavery}
[ link to this | view in chronology ]
Re:
[ link to this | view in chronology ]
Re:
if (defendant!=white) {
guilty();
} else if (defendant==poor) {
guilty();
} else {
innocent();
}
function guilty() {
prison-slavery();
}
function prison-slavery() {
insert code
}
[ link to this | view in chronology ]
Re: Re:
Next you'll be asking how many beta testers ran his code and does he have a printable list of results... :)
[ link to this | view in chronology ]
Re: Re: Re:
[ link to this | view in chronology ]
Re: Re:
[ link to this | view in chronology ]
Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re:
and 2. Pseudocode is usually modelled after something, which is what I asked originally. I modelled my counter-example on PHP.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re:
True.
Pseudocode is usually modelled after something, which is what I asked originally.
Usually, but it needn't be. The original pseudocode was perfectly valid.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re:
2. "The original pseudocode was perfectly valid." -- Smh... you have a weird definition of valid. Not only does this "valid" pseudocode not match any known programming language, but just in reference to itself, it's horribly malformed an illogical... not valid.
For example, what is this garbage;
First off, the single equal sign used in these if-then statements, such as this line, "(defendant)=not_white," aren't comparing, which is two equal signs "==", but is assigning, so that's the first invalidity in this pseudocode, that makes it less pseudocode and more a muddled mess.
Essentially the code as written is saying that "not_white" and "poor" are assigned as the "defendant", and so this code doesn't really testing to see the defendant race and income... I had to make the assumption it was comparing and not assigning.
Secondly, is that a nested if-then or what?
The first if-then, I assumed, was testing the defendant color, and the second if-then tests for income level, but it's ambiguous whether or not the income level test is nested within the color test... and determining nesting is as important as specifying AND versus OR when combining two search criteria.
Third, the final "end if" comment is irrelevant and redundant, because the closing bracket, "}", is the logical end of an if-then statement, and the second closing bracket in an if-then-else statement.
In the end though, the real test of validity is real life... you're being racist and bombastic to claim, either by assignment or by comparison, that all "not_white" or poor people are automatically the defendant and automatically guilty.
There's not one shred of validity, coding-wise or in reality, to any of that quote-unquote, pseudocode.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re:
Again, there is no need for pseudocode to match a programming language.
First off, the single equal sign used in these if-then statements, such as this line, "(defendant)=not_white," aren't comparing, which is two equal signs "==", but is assigning
You are making assumptions based on actual programming languages. Since the conventions of this pseudocode were not specified, it makes more sense to assume conventions that yield a sensible result, such as = meaning comparison.
Secondly, is that a nested if-then or what?
No, it's an if followed by a jump. If the goto executes, the next if isn't evaluated. That's generally what goto means. The "else" and "end if" aren't actually necessary, so "perfectly valid" was maybe an overstatement. But personally I didn't find it confusing, and the quality of the communication of ideas is what's important with pseudocode, not syntactic correctness - since there is no such thing.
In the end though, the real test of validity is real life...
I'll let him defend his own claim if he wants, though I suspect it was more hyperbole than a claim of fact.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re:
I do agree, but it is important for pseudocode to follow logical conventions.
No, not at all. I'm trying to NOT make assumption, which is THE REASON why I asked which language it was based on. Why is this important?
Because many programming languages allow for the use of an assignment operator within the conditions of an if-then statement.
For example...
Not only is the if-then statement testing to see if the query output is null, but it's also assigning whatever values were returned from that query to the variable $results.
The way that Anonymous Coward wrote that pseudocode could very well be interpreted to say one of two things... either...
1. If a person is not white, they're the defendant and the defendant is guilty, or...
2. If a defendant is not white they're guilty.
Now, why is this important? It's important, because it's important to me. I want to understand exactly what that Anonymous Coward meant, and not make any assumptions.
Another Anonymous Coward said it was a joke, but given the nature of anonymous cowards, it's hard to tell if it's the same one or someone else. Either way, the pseudocode doesn't feel like a joke, but instead is making the statement that all minorities and poor people are found guilty, and if you're rich and white, you're not... oh, and that prison is slavery. That's some very racist comments to toss around as a joke.
Oh come now, just admit you have no idea what I'm talking about instead of feigning understanding... the fact is the OP wrote pseudocode that suggests if a person is not white, they'll be found guilty. That's it and no conditional branch for not guilty. The next if-then suggest that if the defendant is poor they're guilty, but this time it gives the "else" statement for innocence.
As you've been saying, I can't take pseudocode literally and I'm not, which is exactly why I asked for clarification.
And given that you keep saying I can't take it literally, I can easily derive two meanings from that malformed attempt at pseudocode, and that is either...
1. All black defendants are found guilty, and all poor defendants are found guilty, or...
2. All poor black defendants are found guilty.
I do believe the writer literally meant #1, and that's how I rewrote my own pseudocode statement.
ROFLMAO... not a programmer, are you?
The only people I know who use pseudocode are programmers, but whatever... the fact is, the only reason to use if-then statements in real life or in pseudocode is for conditional branching, and if you don't think a conditional branch for "innocent" is necessary, then we have a bigger argument than one over the interpretation of someone's pseudocode.
What part of "uses the structural conventions of a programming language" do you not understand about pseudocode. I couldn't care less about syntax, and the proof of that is that I haven't once mentioned closing tags, now have i? But there is conceptual correctness, and that's where my focus is at, so if you're going to invoke the if-then paradigm in trying to explain something, you better fu_king well respect the concepts of nesting, assignments and comparisons when using pseudocode if-then statements.
But on a broader point... if you didn't find anything the OP wrote confusing and you're scolding me for requesting clarification, then I feel it's safe for me to assume you either agree with the OP or have no opinion on what he meant by his pseudocode.
I do have problems with it, but I want to know exactly what the OP meant, so I could better debate the point.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re:
It's clear that you think you are very clever, it's also clear (from this and other posts you've made) that you're a woeful misanthrope who doesn't play well with others.
Perhaps it would be better for us all if you didn't. You Bore.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re:
2. I've already address the usage of the word "mocking".
And 3. as for the rest of you ad hominem... to suggest I somehow dislike humanity is ignorant, and to a large extent, hypocritical, considering you don't trust your fellow human beings enough to not act as an anonymous coward who's running around trolling people.
So really, not only are you a bore, but you're boorish.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Re:
How's that autism working out for you dude?
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re:
Whatever.
I only win if I gain something, but all I've done is waste my time explaining my request for clarification.
Now you can kindly go screw yourself... I won't be responding to you anymore... or any of the Anonymous Cowards on this thread.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Re: Re: Re:
[ link to this | view in chronology ]
Source code have nothing to do here
Since when forensic evidence require such a thing? Now what, for fingerprints to be valid, one must present blueprints of microscope? And how about photos/video? Will police have to present source code for camera? And in tax fraud cases lawyer will require to examine source code of Excel or calculator, because who knows what's in there?
Of cause all this is ridiculous. Tools and software should be checked and certified before use in judgement. But, after certification, tool results are assumed to be valid, unless counterexample is presented. I don't see counterexample here; and except for this DNA sample there must be _other_ supporting evidence.
[ link to this | view in chronology ]
Re: Source code have nothing to do here
If they are solely relying on the magic box, they need to pony up the source code. They need to pony up the logs & evidence the machine was used correctly and maintained properly.
While they might be able to check it with pen and paper, there is a case where they went from the standard 1 in a kajillion chance of it being someone elses DNA to a 1 in 50 chance. Perhaps more evidence would have been required if the jury was told it was 1 in 50 and not 1 in an impossibly large number.
There were those breathalyzers where they got caught with bad code after relying on the hidden code for so long that they went out of their way to block people convicted using the flawed data from demanding their cases be heard again.
[ link to this | view in chronology ]
Re: Re: Source code have nothing to do here
[ link to this | view in chronology ]
Re: Re: Re: Source code have nothing to do here
No, not really. "1 of 1000" is very high (or low, depends how you look) number. You don't convict people in murder on such basis. Other evidence must be presented.
[ link to this | view in chronology ]
Re: Re: Re: Re: Source code have nothing to do here
[ link to this | view in chronology ]
Re: Re: Source code have nothing to do here
No they were not. They were checked by certified equipment, which _does_ produced correct result. Oh, you don't want to check equipment source code, you want one of VW? That's irrelevant. When you check suspension, you are not going to examine blueprints, right?
[ link to this | view in chronology ]
Re: Re: Re: Source code have nothing to do here
So you agree they were checked. So you must be denying that they were certified. So you're saying VW was permitted to sell vehicles that had not been certified to pass emissions testing?
(yes I know this is off topic)
[ link to this | view in chronology ]
Re: Re: Re: Source code have nothing to do here
While there might be more ways to examine decades old DNA evidence, a good easy place to start is the source code. Having someone check code for errors is sadly cheaper than trying to run new DNA tests on small samples that the chain of care and custody could be very questionable of given the wide range.
I have a better car example, everyone swore the brakes always worked that it was operator error. They offered up all of these tests & proof to support it. Then someone finally got the source code and showed that it was possible for it to enter a race situation that did/could (because I am still sure its being litigated somewhere) cause the accidents.
You have someone facing a jury who hear the magic words DNA, Magic DNA machine. They all watch CSI and know that you can get results in 3 minutes and the computers are always right. It is impossible to convince people that what they know is wrong (see also AntiVax, Climate Change, etc etc.) What can raise issues is showing that the machine we've imbued with magical powers can make mistakes.
We all know the examples of flawed CPUs, but how many of us suspect that could happen today... but I bet it has happened again.
We all know that "sometimes" code can be wrong & still generate correct results... how often do we suspect it.
We have heard stories of getting different results based on racial background, but we assume they account for that now... but do we bother to check?
While it might seem silly to you to want to examine the source code, it should be the right of the accused. They didn't even offer up (mind you we have limited information) evidence of any certifications and testings done with the unit. This was well you might build your own and steal our customers as a viable answer to well the magic box said you murdered someone and you can not ask how the magic box got there.
The code prints SSSSS on your boarding pass, but no one gets to ask how the code does that. They are just supposed to accept being treated like a lesser citizen because some coded metrics say you could be a terrorist. Pretty sure we sometimes put to much faith in the code and hold it in to much reverence rather than trust but verify it works as claimed.
[ link to this | view in chronology ]
""TrueAllele, created by Dr. Perlin and in its current version since 2009, is the only computer software system of its kind that interprets DNA evidence using a statistical model. It can single out individuals in a complex DNA mixture by determining how much more probable a match is versus mere coincidence. Complex mixtures can involve multiple people, as well as degraded or small DNA samples. ... "
This could be 1 crackpots math in a black box, and he gets to use trade secrets to keep the ONLY thing doing this from being examined when its using a model of statistics not actual DNA comparison.
http://science.slashdot.org/story/15/10/10/125249/source-code-on-trial-in-dna-matching-ca se
[ link to this | view in chronology ]
Comment on hair analysis
[ link to this | view in chronology ]