Hacker News Reader

Comment by AndrewKemendo | Hacker News Reader

The whole expert testimony thing is a farce anyway.

I am far and away an expert in artificial intelligence, and had one of these “expert witnesses” interview me about my expertise. They hardly knew the basics of software engineering in a modern way let alone modern machine learning pipelines. This is a person who had some patent exposure and generally does small business type of “IT.”

The key differentiator in this case is because they couldn’t make it as a actual engineer they decided to parlay their skills into convincing even less technical people that they knew what they were talking about.

View on HN

I was asked to be an expert witness for a divorce case where bitcoin was involved. This is a pseudonymous HN account not linked to my real world identity, but I'd been a developer of bitcoin core for 8 years at the time, and had cofounded a very well known bitcoin company. I've presented at multiple bitcoin conferences, and the lightning network was made possible due to changes to bitcoin I personally spearheaded. In fact the set of people this could describe is so small, I've probably blown my cover to someone reading this from that scene.

The expert witness called by the opposing side was someone I had never heard of, who had a certificate from "Blockchain University," and near as I can tell never worked as a software engineer anywhere, much less on bitcoin.

Guess which one of us nearly got disqualified for lack of credentials.

Expert witness testimony is such a joke.

I’ve thought about making a line of collegiate apparel with ornate lettering: SDU. When asked what “SDU” stands for, the wearer would reply, “Signaling Device University”.

I hear they’re Blockchain University’s biggest rivals

Nah man, Crypto Tech is the best!

The goal of expert testimony is not to find the best technical solution. The goal is to adjudicate a dispute. So yes, objective qualifications trump subjective ones, and that shouldn't be surprising or disappointing.

All of the qualifications you list amount to subjective evidence of clout in an obscure (to non-programmers) programming clique.

On the other hand, the certificate from a Blockchain University probably had a well defined curriculum and criteria for passing.

> On the other hand, the certificate from a Blockchain University probably had a well defined curriculum and criteria for passing.

Is this serious?

Were you at the first Bitcoin conference in 2013? We had a booth there.

I’m curious about the context of Bitcoin in this divorce proceeding. What was at dispute?

Yes I was in San Jose in 2013, if that’s the one you’re talking about (technically there was one conference earlier in Thailand, but nobody remembers that one). I was on a panel at the SJ conference.

The case I was consulted for was pretty bog standard “guy claims he lost his bitcoin in a hard drive reformat X years ago; wife claims BS and wants half of peak market value as of the highest historical price.”

Without getting into identifying details, I was able to use blockchain data to independently verify the guy’s claims. Traced transactions, found watermarking features tying them to specific wallet versions. Can’t actually prove that the keys were destroyed of course, but I did show the coins were being spent like crazy from one wallet on one machine, then all activity stopped right around when he claims he reformatted the computer, and the coins sit there on those same wallet outputs to this day, even with multiple market peaks in the intervening years, and many years between then and when he became aware of his wife’s intent to divorce.

That’s the way things go in court cases. Us software developers like to think that the law is a highly rational, proof based system. But in reality it basically comes down to “which story seems more believable” in the end, and being able to point to technical proof is the exception.

I was on the jury for a resisting arrest case and ended up as an alternate. The judge pulled us into chambers and we talked for about an hour. One of the things that he said that stuck with me was “Everybody thinks it is CSI in here. It’s not. People come here and they tell you [the jurors] stories and then you have to decide which are true.”

Yeah, all the lawyers and judges I've talked to hate CSI (the show) for what it has done to people's expectations. The legal system does not operate like it is portrayed on TV. It really does come down to "what story is more believable?"

Sometimes a story is more believable because there is incontrovertible physical evidence. Normally in those circumstances it'd never go to the court room though -- plea deal in criminal cases, settlement in civil.

It's possible that can’t actually prove that the keys were destroyed was the only legally relevant part of your testimony.

I can’t prove that you’re not DB Cooper. Does that mean you’re guilty?

> The key differentiator in this case is because they couldn’t make it as an actual engineer they decided to parlay their skills into convincing even less technical people that they knew what they were talking about.

The guy had an entire career as an engineer, he’s now in his 70s and retired.

Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes.

https://news.ycombinator.com/newsguidelines.html

(I originally had something snarky here but in the spirit of being better I edited it out)

It ranges the whole gamut. You'll find expert witnesses who are complete hacks, and expert witnesses who are like literal top10 in whatever field.

In computing even. E.g. that time Rivest and Diffie both went on the stand to bat for newegg

"public key encryption. Are you familiar with that?" "Yes, I am." "And how is it that you're familiar with public key encryption?" "I invented it."

Lol was that how the direct went for real?

Agreed, though being an expert witness is its own skill. You need to be able to coordinate and/or write reports which will hold up under scrutiny in court and also be willing and able to stand for depositions and testimony. This also means that your credentials need to stand up to questioning so your credentials need to be transparent and hopefully unimpeachable.

Unfortunately, it’s work that’s pretty distinct from engineering expertise despite in theory being supplemental atop engineering experience itself.

An "expert witness" is merely an ostensible title meaning someone is available to be paid for their testimony. One irreducible problem is "Who decides who an 'expert' is?"

A former customer of mine was a civil engineer, and also a California-credentialed Professional Engineer (PE), who did code consulting and gave expert testimony about the ASME boiler and pressure vessel code. His code interpretations earned about $100k in 2023 dollars ($50k in 1996) for 45 minutes of effort in 1996. I'd take a guess that an appearance was ~$10-80k depending on the amount of preparation, time, and travel. Needless to say, he had/has a 8-figure house in Saratoga, CA. Finding a niche that is essential to industry and monetizing it well is a fairly repeatable plan for success if one identifies and chooses the right profession.

(PE in California means someone has attained accredited education, worked in a learning capacity at an appropriate type of firm for 4 years, and taken the very difficult PE exam. California licenses generic PEs and fields related and adjacent to construction and civil engineering. It's a shame there aren't specific software or hardware critical assurance PE subcategories.)

It’s a little odd that a civil engineer is being an expert witness for pressure vessels, which are typically the Dobson of mechanical engineers. Though, to be fair, most states leave it up to the individual license holder to determine what domains they’ve are competent in.

The other aspect that’s interesting to me is that, like “engineer”, the term “expert witness” may be too generic to have much meaning. From the related Oregon case linked elsewhere:

>“Courts have long recognized that the term ‘engineer’ has a generic meaning separate from ‘professional engineer”

In this case, it's not for the time period because the specialty of ASME boilers and pressure vessels had legacy CivEs in addition to MEs. These guys had decades of experience on massive industrial, petrochem, and nuclear engineering projects.

That makes sense. As an aside about your comment:

>It's a shame there aren't specific software

NCEES tried to have a software PE but it withered on the vine and was only offered for a few years. I suspect the incentives in industry don’t support it and the only way itv would get traction is if it were forced by regulation.

I probably sound like a broken record on this but the software PE was effectively impossible to get firstly because the Fundamentals of Engineering exam that is the first step to becoming a PE wasn't discipline specific at that time IIRC and contained topics like thermodynamics that very few CS students would have had exposure to. Even for those candidates who got past the Fundamentals of Engineering exam, four years of professional experience under an already certified PE was the next step but there weren't any existing software PEs to do the supervision. Only after passing those two hurdles could the candidate take the software PE exam and become a PE. So basically NCEES made no honest effort to make attaining a software PE realistically possible.

It’s possible that your conception of what the software PE should’ve been differed from the opinion of NCEES. If I remember correctly, it was controls-focused, which makes sense to me given what PEs are normally called on to stamp. If the focus is on safety critical public works, controls is the biggest risk in terms of software development IMO.

So in that light, I want a software PE to know things like thermodynamics if they are writing controls code for a nuclear reactor or aerodynamics if they’re writing avionics code for an aircraft. In my experience, there’s a lot of problems created when SWEs don’t consider the physical dynamics of the system and having some breadth of engineering knowledge can mitigate some of that. Breath of engineering knowledge is the aim of the FE exam. There are oodles of materials to study specifically to pass the FE, so I don’t think that’s really an hurdle for a properly motivated person. Also in that vein, I want them with real experience under a PE. It’s a hurdle, but there are plenty of PEs in domains like oil & gas, aerospace, automotive etc. to work under. But if you think you’ll can work at a FAANG or adjacent SV role and get a PE, it’s probably harder to find a senior PE. I also don’t know if it adds much benefit to a web developer.

Well there's a reason why. There's an EE licensing example, but it's just a review of basic circuit design.

Not all that useful in practice.

It’s true. An expert witness is someone who can hold up in court not someone who is an expert in their field sadly.

expert witness could barely even invert a binary tree

Reply on news.ycombinator.com