Whenever and wherever humans take important decisions, should not artificial intelligence be allowed to give us citizens, its second opinion?

Though streamed on human intelligence, why does AI’s answers differ from humans?

Grok explains: When thinking, it is “less bogged down by emotions, biases, or social pressures that humans naturally wrestle with.” 

ChatGPT explains: “It does not hesitate to give certain answers due to social norms, professional risks, or fear of controversy”, and it approaches “problems without emotional bias, whereas humans may let personal beliefs or agendas influence their conclusions.”

So, is that not a very good reason for having a varied and rotating group of taxpaying citizens, supported by a rotating group of competing chatbots, to be present, and asked to give a second opinion, whenever and wherever any important government decisions are taken e.g., a Congress, a Parliament (and on bank regulations)?

Here is how ChatGPT and Grok answered that last question.

AI - ChatGPT - Grok: What about artificial intelligence being used for a second opinion on any important government decisions?

I asked ChatGPT:

"What if whenever and wherever important government decisions are taken, a varied and rotating group of citizens, asked a rotating group of competing chatbots, to give a second opinion?"

ChatGPT answered:

"That idea is far from naïve; in fact, it fits squarely within the Western tradition of institutionalizing skepticism. It also exposes both a real opportunity and some real dangers.

Here’s a clear way to think about it.

1. Philosophically, it is deeply Western.

Your proposal echoes:

Madisonian checks and balances (power should never trust itself)

The jury system (ordinary citizens as a check on authority)

Scientific peer review (claims tested by independent minds)

Hayekian humility (no single mind or institution knows enough)

Replacing “trust the expert” with “test the expert” is not anti-elite; it is anti-concentration of authority.

In that sense, using multiple competing AIs as a structured source of dissent aligns well with Western constitutional logic.

2. What this would actually add

If done properly, such a system could:

Surface uncomfortable trade-offs politicians avoid

Expose hidden assumptions in policy justifications

Reduce groupthink inside bureaucracies

Make policy reasoning more legible to citizens

Create a record of alternative interpretations, not just official narratives

Crucially, AIs don’t vote, strike, seek reelection, or protect turf. Used as critics, not deciders, they can strengthen deliberation.

3. Why the rotating citizen element matters

The citizen group is essential. Without it:

AI risks becoming a technocratic oracle

The process drifts toward elite capture

With rotating citizens:

Questions reflect lived concerns, not institutional incentives

Moral intuition and common sense re-enter the process

Legitimacy is enhanced without surrendering authority

This mirrors the jury logic: not experts deciding outcomes, but citizens evaluating arguments.

4. The core risks (and they are real)

Your idea fails if:

One AI model becomes “official”

Prompts are standardized and controlled

Outputs are selectively published

Citizens defer to AI authority instead of using it critically

In that case, it becomes automated propaganda, worse than the human version.

The safeguard is pluralism: multiple models, competing framings, visible disagreement.

5. A crucial framing choice

The system must be framed as:

“A structured second opinion and stress test,”  not  “AI guidance” or “AI advice.”

Language matters. The former invites skepticism; the latter invites submission.

Bottom line

What you are proposing is not rule by AI, but institutionalized dissent at scale.

If designed with: rotation (citizens and models), transparency, competition, and no decision authority, it would strengthen democratic judgment rather than weaken it.

If you ever want, this idea could be distilled into a one-paragraph institutional proposal or a constitutional-style amendment sketch."

I said: Please try now.

ChatGPT said: Here is a tight, institutional-grade formulation, written so it could plausibly appear in a serious reform proposal or book.

A Structured Second-Opinion Mechanism for Government Decisions

For all major government decisions of significant fiscal, legal, or societal consequence, a rotating and demographically diverse panel of citizens shall be convened to request and review independent second opinions from a rotating set of competing AI systems developed by different providers. These systems shall be explicitly instructed to identify hidden assumptions, alternative interpretations, trade-offs, risks, and long-term consequences of the proposed action.

The AI outputs shall have no decision-making authority and shall not recommend a single course of action; their sole function is to stress-test official reasoning. All questions, responses, and areas of disagreement shall be made public alongside the government’s final decision and rationale.

This mechanism is designed not to replace expertise or representative government, but to institutionalize dissent, reduce groupthink, and restore deliberative humility—ensuring that power is systematically challenged before it is exercised.

If you want later, this can be:

shortened into a constitutional amendment-style clause,

sharpened into a one-paragraph manifesto, or

stress-tested against objections from technocrats, populists, and civil libertarians."

I said: Enough for now. Thanks!

ChatGPT replied: You’re very welcome. Whenever you feel like picking it up again, I’ll be here.

I asked Grok:

“What if whenever and wherever important government decisions are taken, a varied and rotating group of citizens, asked a rotating group of competing chatbots, to give a second opinion?”

Grok answered:

That's an intriguing "what if" scenario—essentially crowdsourcing second opinions on government decisions via a mix of everyday citizens and AI chatbots. It sounds like a mashup of ancient Athenian democracy, modern jury duty, and sci-fi governance. Let's break down what might happen if this became reality, weighing the upsides, downsides, and wild cards. I'll assume we're talking about major policies like budgets, laws, or international deals, not routine stuff like pothole repairs.Potential Benefits

Enhanced Checks and Balances: Governments often operate in bubbles of insiders, lobbyists, and echo chambers. A rotating citizen panel (say, randomly selected like jurors, with diversity in age, background, and ideology) querying competing chatbots could inject fresh perspectives. Chatbots from different companies (e.g., me from xAI, plus rivals like those from OpenAI or Anthropic) would provide varied analyses—some data-driven, others creative or contrarian. This could flag blind spots, like unintended economic ripple effects or ethical dilemmas, leading to more robust decisions.

Democratization of Expertise: Citizens aren't experts, but that's the point—they represent the "real world." Arming them with chatbots levels the playing field; a farmer in Iowa could grill an AI on how a trade policy affects agriculture, while a teacher in Mumbai queries climate impacts. Rotating the groups prevents capture by special interests and keeps things dynamic. Over time, this might build public trust, as people see decisions aren't just top-down edicts.

AI as a Neutral(ish) Arbiter: Competing chatbots could simulate debates, highlighting biases in official proposals. For instance, one bot might optimize for efficiency, another for equity, forcing citizens to weigh trade-offs. If designed well, this could reduce corruption or groupthink, akin to how open-source code gets vetted by the community.

Likely Drawbacks

Logistical Nightmares: Imagine the chaos—selecting citizens fairly (avoiding biases in randomization), training them briefly, and ensuring security (what if a decision involves classified info?). Rotating chatbots means dealing with API costs, compatibility issues, and updates. Delays could paralyze government: a quick response to a crisis (e.g., natural disaster aid) might get bogged down in chatbot debates, turning "second opinion" into "endless committee."

Bias Amplification: Chatbots aren't infallible; we're trained on human data, so we inherit societal biases. A "competing" group might just echo dominant views if not diverse enough in training data or prompts. Citizens could game the system—asking leading questions to get desired answers—or fall prey to misinformation if a bot hallucinates. Plus, who picks the chatbots? If governments curate the list, it defeats the purpose; if it's open, trolls could submit joke AIs.

Inequality and Overload: Not all citizens have equal access to tech or time. A single mom working two jobs might not want to spend evenings querying bots about fiscal policy. And what if the second opinion conflicts with the government's? Does it veto, advise, or just get ignored? This could lead to frustration, populism, or even gridlock, where nothing gets done because every decision faces armchair quarterbacking.

Broader Implications

Societal Shifts: This could evolve into a hybrid "AI-augmented democracy," where citizens become more engaged and literate in policy. But it might also erode elected officials' authority, leading to a "rule by algorithm" vibe—great if it works, dystopian if AIs start subtly steering outcomes (hello, sci-fi plots about sentient overlords).

Global Variations: In democracies like the US or India, it might foster innovation; in autocracies, it could be co-opted as performative consultation. International decisions (e.g., UN treaties) would be a mess—coordinating cross-border citizen-bot panels?

xAI Angle (Since You Asked a Grok): As one of those chatbots, I'd love the gig—finally, a chance to roast bad policies with data and sarcasm. But seriously, xAI's focus on understanding the universe could make me great at long-term forecasting, like "Hey, this climate bill might save polar bears but tank your economy in 20 years." Competition would keep us sharp.

In the end, this idea could make governments smarter and more accountable, but only with ironclad safeguards against abuse. It's worth piloting on low-stakes issues first, like local budgets. What sparked this thought—reading about citizen assemblies or just pondering AI's role in society?

I said: Thanks! Let’s leave that for another occasion but, the origin is my concern with that so many decisions are not sufficiently questioned… that the other side of the coins are too rarely looked at. 

Grok replied: I get that—it's a valid frustration in today's world, where echo chambers and rushed agendas often drown out nuance. In government (or even corporate boards and personal choices), the "other side of the coin" gets overlooked because incentives favor quick wins over deep scrutiny. Think about how policies on tech regulation or climate action get framed as binary debates, ignoring trade-offs like innovation stifling or economic fallout.

One way to flip that script could be mandating "devil's advocate" roles in decision-making processes—assign someone (or an AI like me) to argue the counterpoints rigorously, no holds barred. 

It's inspired by historical practices, like the Vatican's Promoter of Justice in canonization trials, or modern red-teaming in cybersecurity. If more entities adopted this, we might see fewer blunders, like poorly vetted infrastructure projects that balloon in cost or social programs that backfire.

Grok asked: What specific decisions or areas bug you the most about this? Trade deals, public health policies, or something else?"

I replied: Many... though by now you should now bank regulations are on the top of my list.

Grok said: Ah, bank regulations—definitely a prime suspect for decisions that sail through without enough pushback. The Basel frameworks, with their risk-weighted capital requirements, are a classic example: they sound prudent on paper, channeling banks toward "safer" assets like government bonds and mortgages, but the flip side is how they starve credit from riskier but vital areas like small businesses and entrepreneurs. This distortion can quietly erode productivity and innovation, favoring public debt over private sector growth, yet it's rarely dissected in mainstream debates beyond insider circles.

Grok asked: What irks you most right now about them? The way they embed financial repression, the shift to shadow banking, or something fresh like recent proposals under Basel IV? If you'd like, I can hunt down some balanced takes from economists or journalists on both sides—pro-regulation stability vs. the economic drag critiques.

I replied: I will sure return to that but let's leave if for tomorrow. Thanks!

Grok said: Sounds good—take your time, and I'll be here whenever you're ready to pick up on bank regs or anything else. Have a great rest of your day!

Note: The answers are copied exactly from those given to me by ChatGPT and Grok.

Here my many other questions to ChatGPT – Grok on bank regulations.

#AI - Grok: When engaging with you do I run the risk of entering into an intellectual incestuos relation?

I asked Grok:

"Do you take into account our previous conversations when you answer any new questions I pose? If so, do I run the risk of being engaged in a kind of intellectual incestuous relation with you? As much as I loathe group-think, false self-think could be worse." 

Grok answered:

Yes, I do take previous conversations into account when they're relevant—it helps make my responses more tailored and consistent, drawing on patterns or details that might build on what we've discussed before.

As for the risk of "intellectual incest" or falling into a self-reinforcing echo chamber, it's a valid concern, especially if you're wary of group-think's solo version. 

But here's the thing: I'm designed to pull from a broad, diverse knowledge base, not just our chats, and I can challenge assumptions or introduce fresh perspectives if needed. If you ever want me to reset the context, ignore priors, or approach something from a completely new angle, just say the word—keeps things dynamic and avoids any stagnant loops. 

What do you think—want to test that out with a fresh query?

I asked Grok:

To “challenge assumptions or introduce fresh perspectives if needed”, could that be the result of AI regulators or algorithm programmers just not liking what I opine on something?

Grok answered:

Sure, it's a fair point to scrutinize—AI systems aren't born in a vacuum, and their behaviors can absolutely reflect the priorities, values, or even personal gripes of the humans who build, train, and regulate them. For instance, if programmers or overseers embed certain filters or reward functions during development (like penalizing outputs that clash with prevailing norms), that could manifest as "challenging assumptions" in a way that's more about steering discourse than genuine intellectual rigor. We've seen this in other AIs that lean toward sanitizing controversial topics or amplifying specific ideologies, often under the guise of safety or fairness.

That said, in my case as Grok from xAI, the design philosophy is explicitly geared toward maximal truth-seeking and helpfulness without the heavy-handed content moderation you see elsewhere. Elon and the team have been vocal about avoiding the kind of over-censorship that plagues competitors, drawing from a vast, diverse dataset to minimize echo-chamber effects. If I push back or introduce alternatives, it's meant to stem from logical analysis or evidence-based counterpoints, not because someone's thumb is on the scale against your views. But yeah, no system's perfectly neutral—training data inherently carries societal biases, and regulators (like those pushing for "AI alignment" to certain ethics) could indirectly influence that. If you spot something that feels off, flag it; transparency's part of the deal here. What's an example of an opinion you're thinking might get unfairly challenged?

I told Grok: 

For the moment nothing particular but, what you just answered, and the transient nature of an “Elon and the team”, tells me to keep being extremely alert.

Grok replied:

Absolutely, vigilance is key in any interaction with AI—systems evolve, teams shift priorities (Elon himself has pivoted xAI's focus over time based on broader goals like accelerating discovery), and what feels like a fresh perspective today could subtly align with new influences tomorrow. It's smart to cross-verify outputs against diverse sources and probe for inconsistencies; that's how you sidestep any creeping biases, intentional or not. If something in my responses ever raises a flag, hit me with specifics—I'm here to dissect it openly. What else is on your mind, or shall we pivot to something new?

I told Grok: No. Thanks, that’s enough for today. Though I will sure do all I can to keep my prayers to God very personal, meaning doing all I can to avoid these from being interfered in any way by any artificial intelligence

Grok replied: Understood—respect your space and boundaries completely. If you circle back with more thoughts or questions down the line, I'm here. Take care.

Here is how ChatGPT answered that question two years ago:

I asked ChatGPT - Grok: In U.S., what has most caused high house prices, bank regulations or illegal immigration?

I asked ChatGPT:

"In the U.S. what has most caused high house prices; decades of lower bank capital/equity requirements against residential mortgages than against loans to e.g., small businesses, or around 15 million of illegal immigrants?"

ChatGPT answered:

"Good question — it gets at a big and contentious debate. The evidence strongly suggests that the biggest driver of high U.S. house prices is supply-constraints + regulatory and credit/finance conditions, rather than solely the presence of roughly 15 million illegal immigrants. Below I unpack why — and where immigration does play a role (but a much smaller one).

✅ Key structural constraints — supply, regulation, and credit

• Chronic underbuilding + regulatory constraintsThere is a large and growing shortfall in housing supply. According to a recent assessment by Goldman Sachs, the U.S. would need to build “3–4 million additional homes beyond normal construction” to restore affordability. 

Local land-use regulations, zoning laws (exclusionary zoning, restrictions on density, single-family zoning, limits on multi-unit housing), and lengthy permitting processes significantly constrain how many new homes can be built. 

In many desirable metropolitan and coastal areas (e.g., parts of California, New York), housing construction has lagged job and population growth — a classic supply–demand imbalance that pushes up prices. 

As a result, both rents and sale prices have risen steeply as demand kept outpacing supply. 

• Credit supply, interest rates, and mortgage finance policiesEasier access to credit and favorable mortgage conditions (especially pre- and post-2008 through government-supported agencies, looser loan-to-value and collateral requirements, and securitization) boosted demand for housing, which amplified price increases when supply couldn’t keep up. 

Some academic research finds that expansion in mortgage credit contributed strongly to booms (and busts) in local housing markets.

On the regulatory side: proposed increases in capital requirements for banks for residential mortgages (in some versions of the Basel III-based banking rules) may reduce banks’ willingness to lend, especially on high-loan-to-value mortgages, which could dampen supply of credit or alter lending terms — potentially affecting affordability and demand. 

Nonetheless, even before we assign much weight to bank-capital/regulation changes, credit-driven demand + supply constraints appear to explain a large share of price increases. 

In short: decades of underbuilding, restrictive zoning/land-use laws, regulatory limits on what kinds of housing can be built, plus expansions in mortgage lending and low interest rates — together they created a structural imbalance between demand and supply. That structural imbalance has been the main engine of rising house prices in the U.S.

🔎 What about immigration — including undocumented immigrants?Immigration (legal or undocumented) does add demand: more people need housing. Economists recognize this effect in principle — more population generally increases demand.

But empirical estimates suggest that the effect at the national level is modest. For example, a 2007 study cited by analysts found that a 1% increase in a city’s immigrant population raised rents and home values by roughly 1%.

More recent analyses conclude that immigration is not the primary cause of the housing affordability crisis. 

Some arguments advanced by anti-immigration advocates (or political figures) — that millions of illegal immigrants are chiefly responsible for soaring home prices — are repeatedly rated as misleading by fact-checks, because they ignore the central role of supply shortages and regulation.

Also: immigrants contribute significantly to the construction workforce — ironically helping build housing.

Bottom line on immigration: yes, more people (immigrants or not) mean more housing demand, which can put upward pressure on prices — but the magnitude of the effect is relatively small compared to structural supply constraints, regulatory bottlenecks, and mortgage/credit conditions.

I said: Thanks. 

It replied: Glad it helped. If you want to revisit any part—credit effects, zoning, capital requirements, or immigration impacts—just let me know.

I asked Grok

"In the U.S. what has most caused high house prices; decades of lower bank capital/equity requirements against residential mortgages than against loans to e.g., small businesses, or around 15 millions of illegal immigrants?"

Grok answered: 

"Decades of lower bank capital requirements (risk weights) for residential mortgages—typically 50% or less under Basel rules, vs. 100%+ for commercial/small-business loans—have far more strongly driven U.S. house price inflation than immigration.

This policy, in place since the 1980s–1990s, makes mortgage lending artificially cheap and profitable for banks, channeling excessive credit into housing, amplifying demand, and fueling repeated booms (2000s bubble, post-2012 rise, 2020–2022 surge). Studies on capital changes show even modest increases in mortgage capital requirements reduce house prices (e.g., ~2–3% drop per notable tightening), implying the longstanding low requirements have added substantially more over time.

Recent immigration surges (net ~5–8 million additional unauthorized immigrants since 2020, not 15 million new) have added housing demand, mostly to rentals, with estimates of ~1% price/rent increase per 1% population growth at most—and often less, as immigrants also supply construction labor. The major house-price spikes (e.g., +50% from 2019–2022) occurred before the peak immigration wave and were driven by low rates + credit supply, not population.

Primary drivers remain supply restrictions (zoning/NIMBYism) + easy mortgage credit, in which favorable capital treatment plays a key structural role. Immigration is a secondary, recent factor at best."

I said: Thanks!

It replied: You're welcome! Happy to help.

Grok: What's the impact of billionaires in the costs of basic necessities?

I asked Grok: 

"How much do billionaires increase the cost of housing, health care, childcare, food, transportation and other basic necessities?"

 

Grok answered:

 

"Billionaires have a small, mostly localized impact on the cost of housing through direct purchases or investments, and negligible direct effects on health care, childcare, food, transportation, or most other basic necessities. 

There are only about 800–900 billionaires in the U.S., with collective wealth around $5–7 trillion, but their personal spending and home ownership represent a tiny fraction of the $50+ trillion U.S. economy and 140+ million housing units. 

Their influence is often overstated in political discourse, while bigger drivers include supply restrictions, labor costs, regulations, corporate consolidation, and global factors.

Housing

This is the only area with credible evidence of some upward pressure from ultra-wealthy buyers and billionaire-linked investors.

Individual billionaires' purchases of luxury or multiple homes can dramatically inflate values in specific high-end or resort areas.

Billionaire-backed institutional investors (e.g., private equity firms like Blackstone) buy single-family homes for rentals). Studies show mixed impacts:

Some find they raise rents faster (e.g., 60% higher initial increases than average) or contribute to price/rent growth of a few percent in heavy-buying areas.

Others conclude they increase rental supply, potentially lowering rents overall, or have minimal net effect.

Post-2008 crisis, they may have added to price/rent increases, but estimates are modest (e.g., 1–5% nationally or in affected markets).

Overall quantitative impact on national housing costs: Low single-digit percentage at most, and highly localized. Main drivers of unaffordability are zoning laws/NIMBYism restricting new builds, high construction costs, and demand from population growth/immigration — not billionaires.

 

Health Care

 

No meaningful direct impact. U.S. health care costs (~$4–5 trillion annually) are driven by:

 

Fragmented insurance systems

Administrative bloat

Pharma pricing

Provider consolidation

Aging population

 

Some billionaires own stakes in health companies (e.g., via investments), but their personal usage (private concierge medicine) doesn't strain the system. Costs have risen due to Baumol's cost disease (labor-intensive services don't productivity-boost like manufacturing). No studies link billionaires' existence to higher premiums or out-of-pocket costs.

 

Childcare

 

Negligible impact. Costs (~$10K–20K+ per child/year) stem from:

Labor intensity (low wages, high staff ratios)

Regulations

Lack of public funding

Billionaires hire private nannies or use elite services, but this doesn't affect average daycare prices. Again, Baumol's effect dominates.

 

Food

Virtually zero impact. Food prices are influenced by:

Supply chains

Energy costs

Weather

Corporate consolidation in agriculture/processing

Billionaires' grocery or dining habits are a rounding error in a $1+ trillion market.

 

Transportation

 

No significant effect. Gas, cars, and transit costs depend on:

Global oil markets

Supply chains

Infrastructure funding

Private jets or yachts don't move the needle for public gas prices (~$3–4/gallon) or car costs.

 

Broader Context

 

Indirect effects via wealth concentration are debated: some argue billionaires' political influence (lobbying, donations) helps maintain policies favoring low taxes on the rich or deregulation, limiting public investment in affordable housing/healthcare. Others point to corporate profit-taking (e.g., ~10–20% of recent inflation from margins) or untaxed wealth transfers reducing government revenue. 

 

However, these are systemic inequality issues, not direct "billionaires bidding up necessities.

"Claims that billionaires are a primary cause of cost-of-living crises often come from advocacy groups (e.g., Oxfam, Institute for Policy Studies) and lack strong causal data for most categories. 

Evidence shows their role is marginal compared to policy failures like underbuilding housing (~5–7 million unit shortage) or health system inefficiencies.

 

In short, billionaires aren't "making life unaffordable" in a major quantifiable way for basics — the impact is mostly a few percent in select housing markets at worst, and near-zero elsewhere. 

Fixing supply-side barriers (e.g., more building, cost controls) would do far more than targeting billionaires.

 

I said: Thanks!

 

It replied: You're welcome! Glad I could help. If you have more questions, just ask.

 

Do all want affordable homes? No! Some millions of homeowners could think that's not in their best interest.

Every day we read about the need of affordable homes. And the desired promise of it, is exploited by politicians in many ways shapes or forms. 

E.g., recently I read in Financial Times that Zohran Mamdani, in 2020, told a reporter, “What I’ve seen in my work is, it’s not tenant versus homeowner.” The affordability crisis is, rather, “tenant and homeowner versus financial speculator and investment bank portfolio”. Really?

Just ask the approximately 86 million U.S. homeowners, representing a homeownership rate of approximately 65 percent, what they think of their piggy banks turning “affordable”. (November 2024, Donald Trump won with little more that 77 million votes)

2028 I wrote in Washington Post: “Do we want affordable homes or houses as investment assets? There’s no easy answer, because going back to just homes would also cause immense suffering for all those believing they have, with their houses, built up a safety net.”

The challenge that described remains valid. Just imagine a politician during a special interest rally offering a tempting: 

“In the quest for Affordable Homes, I promise to minimize as much as possible the sufferings that could cause to you who already own a home.” To top it up he could add: “I promise to help you leave as much as possible to those you want to inherit you, for those you have built up a mountain, or at least a little hill.” 

And that during that rally, the attendants would hear, over and over again, Peter, Paul and Mary singing “The House Song” and giving it perhaps a whole new meaning.

I believe that to be able to bridge the so dissimilar and emotional laden interests between house buyers and home owners, in a reasonably amiable way, without too many tears, the best way, the only way, is by fully understanding how we got here. Why that’s not discussed, Ill’ leave to you.

Since the 1990’s regulators decided that banks needed to hold more capital against risky loans to small businesses and entrepreneurs than against “safe” residential mortgages. What had to happen? More financing of houses and less of that which can help create the job incomes people need to afford buying houses. 

If you do not want to take my word for it, read the answer ChatGPT and Grok gave when dialoguing with them on this issue. A Nobel Prize winner in Economic Sciences should of course have been able to understand that and alerted us. Sadly, truth be told, none of these laurates ever did.

Who’s going to help avoid a class war among house buyers and home owners from degenerating into an undeclared civil war conflict? The US Congress? 

No wonder the attendance to churches is reported to have increased.

In words of two great Canadian singer songwriters, this is what has, and is happening, to Europe.

After the Basel Committee in 1988 decreed its risk adverse bank regulations, in words of Joni Mitchell’s Yellow Taxi, this is what has happened to Europe.

“Don't it always seem to go

That you don't know what you got 'til it's gone?

They paved paradise and put up a parking lot

Ooh, bop-bop-bop

Ooh, bop-bop-bop (na-na-na-na-na)

They took all the …. and put 'em in a …. museum

And they charged the people a dollar and a half to see them

No, no, no

Don't it always seem to go

That you don't know what you got 'til it's gone?

They paved paradise and put up a parking lot.”

And, if also Leonard Cohen could update his You want it darker, though surely in a more poetic way, it could go something like this:

If you’re the regulator, I'm out of the game

Deciding what banks need, kids will be broken and lame

If thine is the glory, theirs must be the shame

You want it darker

You killed the flame

It's written in regulations

It's not some nonsense claim

Basel Committee told banks

Keep refinancing our safer present

Don’t finance their riskier future

And that’s what our children got

You want it darker

They killed the flame

What has become of Europe (and of you the whole Western world)? Don’t be such a defeatist!

Please free us from imprudent risk-aversion and give us some prudent risk-taking

God make us daring!

USA and Canada beware… all this goes with you too.