Latest / Deep Dive / Why Software Developers Should Think More Like Plumbers Than Programmers
Transcript
- 0:00Imagine, just for a second, that you're you're stepping onto an
- 0:03elevator in a brand new, like 100 story skyscraper.
- 0:06OK, setting the scene. I like it.
- 0:08Right. So the doors slide shut, you hit
- 0:10the button for the top floor, and you feel that familiar lurch
- 0:14upward. Now imagine finding out in that
- 0:16exact moment that the architects and the engineers who built this
- 0:20tower, that they weren't bound by any building codes
- 0:23whatsoever. Oh.
- 0:24Wow, yeah, that's terrifying. Exactly.
- 0:26There are no universal safety standards, no legal
- 0:30accountability if the cables just snap.
- 0:32And honestly, half of the people who designed the structural
- 0:34supports, they don't even consider themselves professional
- 0:37engineers. They just, you know, really like
- 0:39building things. I mean, you would probably be
- 0:41frantically pressing the emergency stop button at that.
- 0:43.100% because the stakes are just way too high.
- 0:47Yeah. When fiscal safety is on the
- 0:48line, we expect strict universal rules, right?
- 0:52Like rules overseen by a recognized governing body and
- 0:55totally backed by the full weight of the law.
- 0:58We definitely don't leave structure integrity up to the
- 1:00whims of like hobbyists. But, and this is the wild part,
- 1:05we are actually living in that unregulated skyscraper right
- 1:08now. You really are.
- 1:09When you look at the digital infrastructure holding up our
- 1:11modern world, from the banking algorithms that manage your life
- 1:15savings to so software running medical devices to the networks
- 1:19keeping airplanes in the sky, that terrifying elevator ride is
- 1:24basically our daily reality. It's a genuinely unsettling
- 1:27realization when you actually stop and look closely at it.
- 1:30It is, and uncovering how we got here, and you know what happens
- 1:34next is our mission. For today's Deep Dive, we're
- 1:37drawing from a really fascinating collection of essays
- 1:40from philosophers, IT veterans, even a former High Court judge,
- 1:44all compiled under the title Professionalism in the
- 1:46Information and Communication Technology Industry.
- 1:49Yeah, and it forces us to confront this massive blind spot
- 1:52we have. Technology is advancing at
- 1:54breakneck speed, but the rules governing the actual people who
- 1:57build it, they're totally stuck in the past.
- 1:59So stuck. Right.
- 2:00And this deep dive really questions the very foundation of
- 2:04how we hold tech creators accountable.
- 2:06I mean, we have to examine why relying on the government to
- 2:09regulate tech is, practically speaking, impossible.
- 2:12Right, which goes against our instincts.
- 2:14Exactly. We'll also get into why the
- 2:17person fixing your computer or coding your banking app might
- 2:20actually have more in common ethically with a plumber than a
- 2:23doctor. And ultimately, why the
- 2:25safeguard for our digital future isn't a new piece of legislation
- 2:28at all. OK, let's unpack this, starting
- 2:30with, well, the most obvious question.
- 2:33Whenever we see a Wild West scenario like this, the
- 2:35immediate public reaction is just, you know, to demand that
- 2:38the government step. In right, pass a law, fix the
- 2:40problem, yeah. So why can't we just legislate
- 2:43our way out of this? Well, what's fascinating here is
- 2:45the structural mismatch between how the law works and how
- 2:48technology works. I mean, the legislative process
- 2:51is deliberately slow. Painfully slow sometimes.
- 2:54Exactly. But it's designed to be careful,
- 2:55right? It's debated.
- 2:56It's built upon historical precedent, but technology, by
- 3:00its very nature, it just destroys precedent.
- 3:03Yeah, it completely rewrites the rules.
- 3:05Former Australian High Court Justice Michael Kirby
- 3:08contributed to these essays, and he characterizes the law in this
- 3:13space as, quote, marching in the rear and limping a little.
- 3:17Oh, that's such a great visual, limping a little.
- 3:19Which makes total sense because lawmakers are usually what,
- 3:23reacting to a problem that has already caused damage rather
- 3:26than predicting the future, right?
- 3:28And even when they do try to predict the future, the
- 3:31technology just completely outruns them.
- 3:33Kirby actually points to an experience from his own career
- 3:36in the late 1970s. Oh, the medical one.
- 3:39Yeah, he was chairing the Australian Law Reform Commission
- 3:43and they were tasked with drafting rules for human tissue
- 3:46transplants, which at the time was this cutting edge medical
- 3:49and legal frontier. So they were rushing to get this
- 3:52legal framework finished, but before the ink was even dry on
- 3:55their drafts, in vitro fertilization and embryo
- 3:58transplants emerged as like a viable reality.
- 4:02Wow, so while they were still in committee meetings, the science
- 4:05they were supposed to be regulating mutated into
- 4:08something completely different? Exactly.
- 4:10The drafted laws were instantly inadequate.
- 4:13That feels like trying to mandate the size of a steering
- 4:16wheel for a race car, only to look up and realize the
- 4:20engineers have already invented a self driving system that
- 4:23doesn't even have a steering wheel.
- 4:24That is a perfect way to look at it.
- 4:25And, you know, we saw this play out on a massive global scale.
- 4:28Shortly after that, in 1980, the OECD, the organization that
- 4:33helps set international economic policies, they drafted global
- 4:37privacy guidelines. Right, the 1980 privacy rules.
- 4:39Yeah, and they based these rules on a perfectly reasonable
- 4:43logical assumption at the time that data would only be used for
- 4:47the primary purpose for which it was collected.
- 4:50Right, like if a local store collected your home address to
- 4:52ship you a catalog, they just assumed that was the end of the
- 4:55transaction. The whole law was built around
- 4:57that finite utility. And.
- 4:59Then the modern Internet arrived and just turned that entirely
- 5:02upside down. And the entire business model of
- 5:05a search engine or a social media platform is based on
- 5:08collecting data for one, you know, innocuous reason and then
- 5:12monetizing it for 1000 secondary reasons.
- 5:14Yep, building behavioral profiles, targeting ads, selling
- 5:18analytics. The technological leaf
- 5:20completely obliterated the foundational assumption of those
- 5:241980 privacy laws. It's crazy it.
- 5:27Is But Kirby raises an even more chilling paradox about this
- 5:30whole lag, because in the face of this rapid change, lawmakers
- 5:35often just freeze. But inaction is actually a
- 5:39decision. Wait, how do you mean?
- 5:41Well, if a government fails to pass a law regulating something
- 5:44complex, say human cloning or advanced AI, just because the
- 5:47science is moving too fast for them to understand, their
- 5:50silence is effectively A deliberate decision to allow it
- 5:53to continue unchecked. Right.
- 5:55Choosing not to decide is still making a choice, but OK.
- 5:58If the law is always limping behind and government silence
- 6:01basically acts as a green light, aren't we just handing total
- 6:05authority over to like transnational tech corporations?
- 6:08Well, it doesn't this mean we're just letting them dictate our
- 6:10rights through the software they build?
- 6:12If we connect this to the bigger picture, yeah, that fear is the
- 6:16exact reality we are living in right now.
- 6:18It brings to life Lawrence Lessig's famous concept that
- 6:21code is law. Code is law.
- 6:24That's a heavy phrase. It is because the software
- 6:26architecture itself dictates what you can and cannot do,
- 6:29completely bypassing democratic lawmaking.
- 6:32They actually analyzed the Stevens V Sony PlayStation case
- 6:36in the book to illustrate this. Oh right, with the region
- 6:38locking. Exactly.
- 6:40Sony didn't want people playing unauthorized or pirated games on
- 6:43their consoles. Now they could have relied on
- 6:46the slow Gears of international copyright law to sue every
- 6:50single offender globally. Which would take, what, decades?
- 6:53Yeah, and millions of dollars. Right, so instead they just used
- 6:57the physical code and hardware of the PlayStation console to
- 7:00enforce their regional copyright rules.
- 7:02So they hard coded the law into the machine itself.
- 7:05Like if the CD didn't have the right digital signature, the
- 7:08machine just wouldn't read it. The code literally acted as the
- 7:12police officer, the judge, and the jury right inside your
- 7:15living. Room.
- 7:15Precisely. And that is the core failure of
- 7:17external regulation. If the laws written by
- 7:20governments are too slow and corporate code becomes the de
- 7:22facto law of the land, our only real safety net is the internal
- 7:26regulation of the tech industry itself.
- 7:28The internal regulation, yeah. If the police can't catch the
- 7:31speeding cars because the cars move at the speed of light, we
- 7:34have to make absolutely sure we can trust the drivers.
- 7:38We have to rely on their professionalism.
- 7:40OK, here's where it gets really interesting, because if we are
- 7:44forced to rely on the quote UN quote professionalism of the IT
- 7:48industry, we have a massive problem on our hands.
- 7:52I really do. Like, are they even a
- 7:53profession? We throw that word around a lot.
- 7:56But John Ridge, he's a former president of the Australian
- 7:58Computer Society. He provides this statistic in
- 8:01the book that just completely shatters that illusion.
- 8:03The 6.5% stat. Yes, only about 6.5% of
- 8:08information and communication technology practitioners in
- 8:11Australia actually belong to the Australian Computer Society.
- 8:146.5%. It's wild.
- 8:16And if you compare that 6.5% to doctors or lawyers or, you know,
- 8:20civil engineers in those fields, membership in a professional
- 8:24body isn't just some optional networking club.
- 8:26It is near 100% and crucially, it is tied to the legal right to
- 8:30practice. Right.
- 8:31If you violate the ethics of the medical board, you lose your
- 8:34license. You can no longer legally
- 8:36operate on a patient. Exactly.
- 8:38But in IT, there is no license to lose if a coder builds a
- 8:42deeply unethical algorithm. They might get fired from one
- 8:45company, sure, but they can just walk across the street and get
- 8:48hired by another company the very next day.
- 8:50Which creates this incredibly fragmented, wildly inconsistent
- 8:54culture. The sources actually feature an
- 8:57essay by Clive Botton and he outlines the types of
- 9:00personalities that this lack of structure attracts.
- 9:03Oh his typologies are so spot on.
- 9:05They really are. He talks about the highly
- 9:08broadly technical type. So this is like the brilliant
- 9:12cowboy coder who loves a massive complex challenge, builds an
- 9:16absolute masterpiece of a system, but possesses 0
- 9:19diplomacy, refuses to document their work, and basically
- 9:22terrifies management because literally no one else
- 9:25understands what they just built.
- 9:26We've all worked with that person.
- 9:28And then on the other end of the spectrum, Botton identifies the
- 9:30big picture, narrowly technical type.
- 9:33This is the practitioner who, say, learned one specific
- 9:37software framework A decade ago, and now they just try to force
- 9:41every single complex problem they encounter into that exact
- 9:44same limited solution. Right, Regardless of whether it
- 9:48actually fits the client's needs at all.
- 9:50It's the old saying, if the only tool you have is a hammer, every
- 9:53problem looks like an. Ant exactly.
- 9:55But how do these isolated, you know, quirky personalities
- 9:59actually prevent it from becoming a true profession?
- 10:02Well, because they prevent the formation of a universally
- 10:06agreed upon body of knowledge, or what they call a bupo.
- 10:10Yeah, true professions have a shared baseline, like every
- 10:13doctor understands basic human anatomy, every lawyer
- 10:16understands contract law. Now software engineering, which
- 10:20is a specific subset of tech, they do have the Sweebuck, A
- 10:23defined body of knowledge. OK, so software engineering has
- 10:26one. But information and
- 10:28communication technology as a broad field, it is completely
- 10:31fractured. You have these brilliant rogue
- 10:33operators and rigid 1 trick ponies, but you have no shared
- 10:37foundation that every single practitioner is required to know
- 10:40and adhere to. And that is why Botton concludes
- 10:43that right now IT is merely a semi profession.
- 10:47A semi profession. But let me play devil's advocate
- 10:49here. Does slapping the label of
- 10:51profession on IT actually solve anything or is that just a form
- 10:56of like elitist gatekeeping designed to keep brilliant
- 11:00self-taught coders out of the industry?
- 11:03I mean, if someone can code circles around a university
- 11:06graduate, why should they be locked out just because they
- 11:08don't have a piece of paper? That's a fair pushback, and it
- 11:11raises an important question about what a profession actually
- 11:13does. It isn't just about gatekeeping,
- 11:15right? A true profession is defined by
- 11:18a norm of altruism and a degree of authority over clients.
- 11:22Authority over clients. Yeah, think about a doctor.
- 11:24A doctor tells a patient what they need for their long term
- 11:27health. Even if it's not the easy pill
- 11:29the patient wants today, the doctor has the structural
- 11:32autonomy to say no. Right, they can push back.
- 11:35Exactly. But in IT, tech workers almost
- 11:38always answer to corporate managers and profit margins, not
- 11:42to some ethical board. They're building what the
- 11:44corporation demands. They rarely have the autonomy to
- 11:47say no. This data collection method
- 11:50violates the ethical standards of my field and I refuse to
- 11:53build it. Wow, Yeah.
- 11:55So if it isn't a traditional professional like law or
- 11:58medicine, and the practitioners don't have the authority to act
- 12:01like doctors, what are they? Well, this brings us to a really
- 12:06radical reframing by philosophers John Weckert and
- 12:09Douglas Adini. I love to this part of the.
- 12:11It's brilliant in its simplicity, right?
- 12:13They argue that maybe it doesn't need to look at doctors at all.
- 12:16Maybe it just needs to look at plumbers.
- 12:18Plumbers. It sounds almost insulting at
- 12:20first. It does, but their logic is
- 12:22airtight when you break down the actual criteria for a profession
- 12:25like what makes a professional. You need esoteric knowledge that
- 12:28the general public doesn't possess, you need autonomy and
- 12:32how you execute your work, You need to serve a vital social
- 12:35function, and you need a shared code of practice.
- 12:39Well, a plumber meets every single one of those criteria.
- 12:42They understand complex fluid dynamics and pressure systems
- 12:46that the average homeowner just decent.
- 12:48You don't stand over the plumber telling them which wrench to
- 12:51use. Exactly.
- 12:52They have autonomy and they protect the health and safety of
- 12:55the community by ensuring sanitation, and they operate
- 12:58under strict building codes. The only real difference between
- 13:02a plumber and a traditional white collar professional is
- 13:06that plumbers get dirty. So what matters isn't the
- 13:09prestige of the title, it is the professional attitude.
- 13:12Yes, the attitude. It's taking pride in the work,
- 13:14understanding its impact on the safety of the community, and,
- 13:17crucially, taking responsibility for the outcome.
- 13:21And the stark reality is that the tech industry aggressively
- 13:24avoids that responsibility. They run from it.
- 13:26Literally the text looks at standard software agreements,
- 13:30specifically pointing out an Apple Software End User License
- 13:33Agreement in EULA. It explicitly states that the
- 13:36software is provided quote as is with all faults and with 0
- 13:41guarantee that it will be error free or uninterrupted.
- 13:44I mean, Can you imagine a plumber operating under an AS IS
- 13:47clause? Right.
- 13:48Hey, the pipes are as is. Good luck.
- 13:50Exactly, if they install a pipe and it immediately bursts and
- 13:55floods your living room, they are legally and professional
- 13:57liable to fix it. But if a software company pushes
- 14:01A mandatory update that wipes out your entire business
- 14:04database, they just point to A50 page legal disclaimer you
- 14:08clicked agree on while you were setting up the device.
- 14:10It's insane. Why does the person fixing my
- 14:14toilet have exponentially more accountability than the person
- 14:18building the digital infrastructure that holds my
- 14:19mortgage and my medical records? Well, the lack of accountability
- 14:23stems from a deep psychological mechanism that Neville Holmes
- 14:26talks about. He calls Is it anthropopathism?
- 14:28Anthropopathism. That's a mouthful.
- 14:30It is, but it basically just means the uniquely human
- 14:33tendency to attribute human qualities to machines.
- 14:36Think about the language we use. Every single day we say oh the
- 14:39computer made an error, or the algorithm is biased, or the
- 14:42system rejected my application. Right.
- 14:44We use the machine as a scapegoat.
- 14:46We literally blame the metal and the electricity instead of the
- 14:49people. Exactly.
- 14:50And Holmes despises the term artificial intelligence for this
- 14:53very reason. Really.
- 14:54Yeah, he prefers the term algoristics because calling it
- 14:58intelligence allows us to dodge the uncomfortable truth.
- 15:01Human beings wrote that flawed code.
- 15:04When we blame the algorithm, we successfully absolve the
- 15:07creator. We artificially separate the
- 15:10digital tool from the human maker.
- 15:12So. We have to shatter that
- 15:13illusion. We have to.
- 15:14We have to remember the many hands problem in software
- 15:18development. Just because a banking system
- 15:20was built by a massive team of say, 300 developers broken into
- 15:24tiny silos across 5 different time zones, that doesn't mean no
- 15:28one is morally responsible. A diffusion of Labor cannot mean
- 15:32a diffusion of ethics. But let's, you know, let's get
- 15:34into the trenches here. Let's look at the reality of how
- 15:37these digital projects are actually built day-to-day.
- 15:40Because if the laws fail to keep pace, and if the professional
- 15:43bodies lack the teeth to enforce, rules and legal
- 15:46disclaimers shield the massive companies from liability.
- 15:49It's a perfect storm. Right.
- 15:51So how do we actually guarantee ethical behavior from the
- 15:55individual coder sitting at their desk?
- 15:57Mark Hottie outlines a concept in the book called The
- 16:00Heisenberg Uncertainty Principle of IT projects, and honestly, it
- 16:05explains so much about why tech goes wrong.
- 16:08It's such a brilliant framework. Holly observes that there are
- 16:11three moving targets in any tech project.
- 16:14You have the specifications, So what the software is actually
- 16:16supposed to do, you have the time when it must be delivered,
- 16:19and you have the treasure how many resources and staff are
- 16:23allocated to it. Specs, time and treasure.
- 16:25OK, his Heisenberg principle states that if management firmly
- 16:28locks down one of those targets, the other two will inevitably
- 16:32distort. OK so give me an example.
- 16:34If the boss walks in and says this new app feature absolutely
- 16:37must launch by Black Friday and the date is non negotiable, what
- 16:41happens? Well, the developer is instantly
- 16:43placed in an impossible situation.
- 16:45They know they can't build it safely in that exact time frame,
- 16:48so because the time is locked, the specifications distort.
- 16:51They have to to accrue what's called technical debt.
- 16:53They start reusing old clunky code just to save hours.
- 16:58They skip the rigorous security testing.
- 17:00They hide minor flaws and just hope no one notices until after
- 17:04the launch. Yes, they just have to hit the
- 17:05date. Exactly.
- 17:06The immense pressure of the corporate environment creates a
- 17:09breeding ground for unethical shortcuts.
- 17:11So how do we fix that? We can't just write a rule that
- 17:14says don't take shortcuts. People will always find a way
- 17:17around it, right? And Richard Volkman offers a
- 17:20solution that reaches all the way back to ancient philosophy,
- 17:23actually, virtue ethics. Virtue Ethics.
- 17:25Yeah. Volkmann argues that in our
- 17:27fast-paced tech economy, traditional ethical frameworks
- 17:31just breakdown. Like you can't use
- 17:33utilitarianism which asks you to calculate the best outcome for
- 17:37the most people because in tech the variables change every 5
- 17:40minutes. Well, you can't calculate the 10
- 17:42year outcome of an algorithm that might mutate tomorrow.
- 17:45Exactly, and you can't rely on strict deontology either, which
- 17:48is the ethics of rigid rules, because rigid rules just
- 17:52encourage lawyerly loophole finding.
- 17:54Oh for sure, if you manage mandate that a company can't
- 17:57track a users IP address, they will just invent a new way to
- 18:01track their mouse movements instead.
- 18:02Exactly. They'll always find the
- 18:03loophole. It's like relying on a GPS map.
- 18:06The second a new road is built or a bridge washes out, your GPS
- 18:10is going to drive you straight into a lake.
- 18:12But virtue ethics? Virtue ethics is like having a
- 18:15compass. It always points true north no
- 18:18matter how rapidly the landscape changes around you.
- 18:21I love that virtue ethics doesn't ask what is the specific
- 18:24rule here, It asks what makes a person of good character.
- 18:27It relies on cultivating the inherent culture of being a
- 18:30quote UN quote good geek. Someone who values integrity,
- 18:34honesty, and good judgment above just getting the job done fast.
- 18:38So it's about internalizing those professional standards so
- 18:42deeply that they guide your actions even when the boss isn't
- 18:45looking. Even when there's no specific
- 18:47law forbidding what you were about to do.
- 18:50Yes, and Neville Holmes ties this back to what the highest
- 18:53goal of technology should actually be.
- 18:55He argues we shouldn't just be building tech to drive mindless
- 18:58consumerism or to automate away unskilled labor just to save
- 19:02corporate dollars. Which feels like all we do right
- 19:04now. Right, but we should be building
- 19:06it for noble purposes, like what he calls automatic drill in
- 19:10education. Using the infinite patience of a
- 19:12computer to teach foundational skills.
- 19:15Tailoring the pace perfectly to individual students so human
- 19:18teachers can focus on higher level mentoring.
- 19:20That is the kind of system a truly virtuous IT profession
- 19:23would prioritize. Exactly.
- 19:25I mean, I I love the compass analogy and the vision of tech
- 19:28elevating education is beautiful, but I have to push
- 19:31back. Here is virtue, this internal,
- 19:35personal sense of being a good geek.
- 19:37Is that really strong enough to survive the real world?
- 19:40It's a tough question. Like when you have a manager
- 19:43screaming that a project might ship by Friday, or the
- 19:46department loses its funding and your mortgage is on the line,
- 19:49does an internal compass really stop you from cutting corners?
- 19:53Look, it is the ultimate challenge of the modern
- 19:55workplace, but Volkmann and the other contributors would argue
- 19:59that a strong, shared professional culture is the only
- 20:03thing that can withstand that pressure.
- 20:05The only thing? Yeah, because when the external
- 20:07rules are blurry, your character is quite literally the only
- 20:11safety net left. If the IT profession as a whole
- 20:15shares those virtues, it empowers the individual.
- 20:18It allows the coder to stand up to the manager and say no.
- 20:22Skipping the security protocol violates the core ethics of my
- 20:25profession, and I won't do it. Because they have the backing of
- 20:29their entire fields culture. Exactly.
- 20:31Without that shared culture of virtue, the isolated individual
- 20:35will almost always cave to the corporate demand.
- 20:38Man, we have covered incredible ground today.
- 20:40We started in the halls of the High Court, realized using that
- 20:43the law is moving way too slowly to catch the technology we're
- 20:46building. We really went everywhere.
- 20:48We did. We explored the fragmented Wild
- 20:50West of the IT industry and realized they lack a unified
- 20:53foundation. We compared software engineers
- 20:55to plumbers and discovered the plumbers are the ones actually
- 20:58taking accountability for their work.
- 21:00And finally, we arrived at virtue ethics, the realization
- 21:04that the character of the coder is our ultimate firewall against
- 21:08disaster. So what does this all mean?
- 21:11It means that every single time you download a new application,
- 21:14every time you click agree on a massive end user license
- 21:18agreement without reading it, every time you trust a digital
- 21:21system with your private communications or your finances,
- 21:24you are placing immense, unquantifiable trust in the
- 21:27internal character of anonymous coders you will never meet.
- 21:31Wow. You are betting your digital
- 21:32life that their compass points true north.
- 21:35That is a terrifying and profound thought, and I want to
- 21:38leave you with a final thought to Mull over today, Something
- 21:40that builds on this idea of accountability.
- 21:42We talked about how we use anthropopathism to blame the
- 21:46machine and absolve the human creator.
- 21:49But look at the cutting edge of tech right now.
- 21:51Oh, here we go. We are actively moving toward AI
- 21:54systems that can independently write, test, and execute their
- 21:57own code. So what happens when the human
- 21:59software engineer is entirely removed from that immediate
- 22:02process? If human virtue and a personal
- 22:04compass are our only real safety Nets today, how exactly do we
- 22:08teach a machine to possess character tomorrow?