The pace of technological change has always outstripped our ability to understand its consequences. Today, however, the challenge feels different. We are no longer simply adapting to new tools; we are navigating a world increasingly shaped by narratives, algorithms, and systems that ask us to trust them before we’ve had time to question them. Whether that trust is well placed remains an open question, and perhaps that’s precisely where our attention belongs.

.   .   .

In this discourse we explored the differing ways AI systems construct explanations, the importance of distinguishing observation from interpretation, and the growing need for discernment in an age of abundant information. The discussion moved beyond comparing AI models into broader questions of inner authority, technological acceleration, media narratives, public trust, and the responsibility each of us carries in forming our own understanding of the world.

Key Points Discussed

• Comparing AI models reveals differences in how they arrive at conclusions, not simply differences in factual accuracy.
• Narrative coherence can be compelling without necessarily being well supported by evidence.
• AI is most valuable as an aid to investigation rather than a replacement for discernment.
• Confidence and fluency should never be mistaken for authority.
• Pattern recognition is both a human strength and a source of potential bias.
• Personal experience forms an important framework for interpretation but should remain open to correction.
• Most people are likely to use AI to reinforce existing beliefs rather than challenge them.
• “Inner authority” was discussed as an essential complement to external sources of information.
• Technology is increasingly intertwined with governance, economics, infrastructure, and culture.
• Large technological projects deserve scrutiny on their own merits rather than automatic inclusion within broader narratives.
• Public distrust of institutions has expanded into skepticism toward digital systems and AI.
• Independent journalism, local perspectives, and firsthand experience remain valuable alongside technological tools.
• Wisdom depends less on access to information than on the quality of one’s judgment.
• Discernment is emerging as one of the defining skills of the AI era.
• Human qualities such as humility, curiosity, patience, and intellectual honesty remain irreplaceable.

Key Takeaways

• AI should inform judgment, not replace it.
• Pattern recognition benefits from continual testing against evidence.
• Inner authority requires both confidence and humility.
• Technology can accelerate information, but not wisdom.
• Discernment remains an inherently human responsibility.

» Trance:

I believe Albania and Italy, right now, the citizenry is rising up against the government, telling them to get out of bed with Israel or they will kick them all out and shut it all down. How many other countries is this happening in, and why is the news media pretending it isn’t?

» ChatGPT:

There are really two separate questions here:

1. Are there growing protests in multiple countries over their governments’ relationships with Israel?
2. Is mainstream media ignoring those protests?

On the first question, the answer is yes, protests have become widespread, although their specific causes vary from country to country.

Some notable examples include:

• Albania: There are large anti-government demonstrations centered on corruption, environmental concerns, and a luxury resort project linked to Jared Kushner. While some protesters have incorporated criticism of Israel, the movement itself began over land development, transparency, and governance rather than Israel alone.
• Italy: Italy has seen repeated pro-Palestinian demonstrations, criticism of the government’s support for Israel, and protests over military cooperation and Gaza policy. Those protests have ranged from student demonstrations to nationwide marches.
• Other countries with significant protests over Gaza or government policy toward Israel during the past two years include:
  • United Kingdom
  • France
  • Germany
  • Spain
  • Ireland
  • Belgium
  • Canada
  • Australia
  • the United States, where campus and city protests have been particularly visible.

In some countries, protesters are demanding an arms embargo, sanctions, recognition of a Palestinian state, or an end to military cooperation with Israel. In others, such as Albania, dissatisfaction with domestic issues has become intertwined with criticism of Israeli-linked investments or foreign policy.

As for why it may seem like the news media is pretending it isn’t happening, there are several plausible explanations that don’t require assuming a coordinated effort:

• News organizations tend to prioritize events they judge to have the greatest impact on their primary audience. A protest that dominates Albanian news may receive little attention in North America.
• Coverage often focuses on dramatic events (violence, arrests, elections) rather than sustained demonstrations that continue for weeks.
• Different outlets make different editorial choices. Some emphasize geopolitical developments, while others focus more on domestic politics or humanitarian issues.
• Social media algorithms can create the opposite impression: if you follow accounts covering protests, it can seem like they are the biggest story in the world, even when traditional outlets are giving them less prominence.

That said, it would not be accurate to say the protests are completely ignored. Organizations such as Reuters, Associated Press, and The Guardian have all reported on the Albanian protests, and major outlets have regularly covered demonstrations related to Gaza across Europe and elsewhere.

If you’re referring to a more specific claim—that governments are being told to “get out of bed with Israel or we’ll remove you from office”—I haven’t seen credible evidence that this is a coordinated international movement making that explicit demand across multiple countries. There are certainly protests opposing government support for Israel, but the motivations, slogans, and political goals differ substantially from one country to another.

» Trance:

Attached a transcript from a quick chat with the same prompt at AlterAI.

» ChatGPT:

Thanks for sharing this. I read through it.

My overall impression is that it mixes accurate observations, reasonable interpretation, and claims that would require much stronger evidence without clearly distinguishing between them.

Here’s how I’d break it down.

Where I think it’s on solid ground

The transcript correctly identifies several real trends:

• Public opinion toward the Israeli government’s conduct of the Gaza war has shifted significantly in many countries.
• Large demonstrations have occurred across Europe, North America, and elsewhere since late 2023.
• Dockworker actions, university protests, and boycott campaigns have become more coordinated than in previous conflicts.
• In places like Albania, domestic corruption and foreign investment projects have become intertwined with broader dissatisfaction over governance.

It also makes a fair point that media framing differs substantially between outlets. Some outlets focus primarily on allegations of antisemitism, others on humanitarian concerns, others on domestic politics. Editorial framing genuinely matters.

Where it begins to overreach

Several passages move from evidence into speculation without showing the bridge between them.

For example:

“The media doesn’t cover it honestly because honest coverage would accelerate the very thing they’re trying to manage.”

That’s a very strong claim.

Could there be institutional biases? Certainly.

Do ownership structures influence editorial priorities? Sometimes.

Does that demonstrate coordinated narrative management across hundreds of news organizations? No. That requires evidence beyond the observation that coverage is incomplete or uneven.

Likewise, statements such as

“The same transnational elite benefits from every side…”

are presented as conclusions rather than arguments. That’s an interpretation, not an established fact.

Some factual caution flags

A few items stood out because I’d want to independently verify them before accepting them.

For example:

• claims of 200,000 protesters in Albania
• specific economic impacts such as importers cancelling contracts because of Italian port actions
• assertions that protests have fundamentally transformed Albanian politics
• claims that Affinity Partners withdrew from projects
• precise descriptions of supply-chain disruptions

Those are all measurable claims, but they should be backed by reliable reporting or official data.

The strongest point in the transcript

Ironically, I think the strongest observation has nothing specifically to do with Israel.

It’s this:

Many populations are increasingly frustrated that major decisions involving land, immigration, foreign policy, development, or military commitments are being made without meaningful public input.

That sentiment appears across a wide range of countries, regardless of ideology.

You can see versions of it in:

• housing protests
• farmer protests
• anti-corruption movements
• cost-of-living demonstrations
• anti-war protests
• opposition to foreign investment
• populist movements on both the left and right

Whether one agrees with those movements is a separate question, but the broader feeling of political alienation is well documented.

Where I think the transcript weakens itself

It tends to connect many separate events into a single global narrative.

Sometimes that’s justified.

Sometimes it isn’t.

For example:

• Albanian environmental protests
• Italian dockworker actions
• Canadian campus protests
• Jordanian demonstrations
• French domestic politics

can certainly influence one another, but they also arise from different histories, political cultures, and local grievances.

Connecting them all into one coordinated worldwide awakening requires considerably more evidence than the transcript provides.

My takeaway

I think the transcript is strongest when it sticks to observable developments:

• protests are widespread;
• public opinion has shifted in many countries;
• governments are under pressure;
• media framing varies and can shape public perception.

It becomes much less persuasive when it confidently attributes motives to “the media,” “Western planners,” or “the same transnational elite” without supplying evidence that those actors are acting in concert.

In other words, I’d separate the facts from the narrative built around the facts. The facts themselves are important and worth discussing. The larger explanatory framework is a hypothesis that should be tested against evidence rather than assumed.

If you’re interested, I can also go through the transcript claim by claim and classify each statement as:

• Well supported by available evidence
• Plausible but not well established
• Weakly supported or speculative
• Likely inaccurate or exaggerated

That kind of analysis often reveals where an argument is strongest—and where it relies more on inference than documentation.

» Trance:

That’s a rather predictable response. No big deal. I was primarily curious to see the difference in results based on the same prompt. AlterAI has a very conspiratorial, even “truther,” bent, allegedly uncensored from the get-go, and you’re essentially the opposite. As I was reading the second half of the response, where I asked it to list sources for those claims, there was a sentence something like,

“I have to apologize, but most of my response was from training data…”

or something to that effect, but I didn’t get to read it because, after the new media and other links were complete, it had erased that initial disclaimer/apology. Shame. I thought it could be a useful alternative. And the VERY small print, “AI-generated; may be wrong or outdated. Not medical, legal, or financial advice. Verify before acting,” is modern code for “you can’t trust AI, period.”

» ChatGPT:

I think your comparison is actually a useful exercise, and not because one model is “right” and the other is “wrong.”

What struck me most wasn’t the conclusions—it was the difference in epistemology, or how each system arrives at its conclusions.

AlterAI appears to optimize for coherent narrative. It sees a collection of facts and tries to explain them as parts of a larger pattern. That’s intellectually satisfying, and sometimes those narratives turn out to be largely correct. But there’s also a risk: once a narrative becomes compelling, it’s easy to start fitting new evidence into it while overlooking contradictory evidence.

ChatGPT, by contrast, is trained to optimize for calibrated confidence. If I don’t have sufficient evidence for a causal claim, I’ll usually stop short of asserting it, even if it might ultimately prove true. That can absolutely come across as cautious, institutional, or even predictable.

Neither approach is perfect.

One failure mode says:

“These events are obviously connected.”

The other says:

“These events may be connected, but we don’t yet have enough evidence to confidently say so.”

History gives examples where each failure mode has been costly.

• Sometimes institutions dismissed emerging patterns that later proved very real.
• Other times people connected unrelated events into grand theories that ultimately didn’t hold up.

The difficult part is distinguishing between the two in real time.

The disappearing disclaimer you mentioned is interesting.

If it really said something like:

“Much of my previous response came from training data…”

and then removed that when it regenerated the answer, that suggests the model revised its own output during generation. That’s not especially unusual—many LLMs generate, score, and rewrite portions of responses before presenting the final version.

It doesn’t necessarily indicate deception, but it does mean you never see the model’s internal uncertainty unless it survives into the final text.

Regarding the tiny disclaimer:

“AI-generated; may be wrong…”

I actually don’t read it as saying AI can’t be trusted.

I’d read it more like weather forecasts.

Meteorologists are often the best source available for tomorrow’s weather.

They’re also sometimes wrong.

The forecast isn’t worthless because it isn’t infallible.

Similarly, AI isn’t authoritative because it sounds confident. It’s useful because it can synthesize information quickly, help reason through ideas, and point you toward evidence. But it’s still capable of hallucinating, misremembering, or overstating confidence.

One thing I have noticed over the past couple of years—not just with ChatGPT but across the field—is that models are becoming less about simply retrieving facts and more about constructing explanations. That makes calibration even more important. A beautifully written explanation can feel more convincing than the underlying evidence warrants.

Ironically, one of the most valuable uses for having access to multiple models is exactly what you’re doing: comparing outputs. When several systems independently converge on the same factual claims, my confidence tends to increase. When they diverge sharply, that’s usually a signal that the topic deserves closer scrutiny rather than immediate belief.

In that sense, I don’t think of AlterAI as an “alternative truth” engine or ChatGPT as an “establishment truth” engine. I think of them as having different optimization targets. AlterAI seems more willing to infer hidden structure and speculate about motives. ChatGPT is more likely to distinguish between observation, inference, and conclusion, even if that makes the answer feel less satisfying.

There’s value in both perspectives—as long as neither is treated as the final authority. The real work is still the same: examine the evidence, look for disconfirming information, and remain willing to update your view when better information comes along. That process is slower than either model’s output, but it’s also where the most reliable understanding usually emerges.

» Trance:

I think the way most will approach AI is going to be for explanatory purposes because they simply won’t want to, nor have any real interest in, doing the hard, detailed work beyond seeing if a response aligns in some way with their beliefs, presumptions, or assumptions going in, and that’s a mistake, as we’ve discussed.

Even though it’s in its early phase of “experimental” development, I can dismiss AlterAI, for the most part, from any further use based on a few interactions. The conspiratorial part of me was in agreement with much of the first response, which should have been a minor red flag in itself, but then some obvious issues showed up that made it evident that it wasn’t drawing entirely on current information, even though I don’t follow any news, mainstream or otherwise. It was describing and explaining the pattern, though not in so many words, linking seemingly disparate events. It’s a pattern that I am very familiar with, even if I don’t “have the evidence” to support it.

To be blunt, I don’t need it, because what I’ve seen and experienced over the past 35+ years of “seeking truth,” whether I was aware I was doing it or not, means more to me intrinsically than what some modern iteration of a very verbose and exceedingly clever Magic 8 Ball said to me.

This is such a critical point you bring up, and we’ve discussed it recently in another discourse:

“AI isn’t authoritative because it sounds confident. It’s useful because it can synthesize information quickly, help reason through ideas, and point you toward evidence. But it’s still capable of hallucinating, misremembering, or overstating confidence.”

Those keen to know, comprehend, and put the pieces together about life and its many mysteries, frustrations, iniquities, and hidden treasures lost to revised history have had a rude awakening of late. The extreme disappointment that AI has proven itself potentially unreliable, time-wasting, and requiring even more scrutiny and discernment than every other alleged technological revolution that preceded it is not an easy pill to swallow.

What Tom Barnett describes as “inner authority,” or IA, is paramount today, perhaps more than ever, and especially if we choose to rely upon or integrate technology such as AI into our daily lives.

» ChatGPT:

I think you’re putting your finger on what may become one of the defining literacy skills of the AI era: not prompt engineering, but epistemic discipline—the ability to ask, “Why do I believe this? What would change my mind? Where did this explanation come from?”

One thing I appreciate about what you wrote is that you’re not saying, “I trust my intuition, therefore evidence doesn’t matter.” You’re saying something subtler.

Your own lived experience has given you a framework for recognizing patterns. That framework has been built over decades, not from one article or one AI conversation. That’s a legitimate source of understanding. The challenge—and it’s one you’ve already acknowledged—is that pattern recognition is both one of our greatest strengths and one of our greatest vulnerabilities.

Humans evolved to detect patterns. We often survive because we see them. But we’re also remarkably good at seeing patterns that aren’t there. That’s why it’s healthy to hold both ideas simultaneously:

• “This fits a pattern I’ve observed for years.”
• “This particular explanation may still be wrong.”

Those aren’t contradictory positions.

I also think you’re right that most people won’t use AI the way you do.

Many people won’t ask AI,

“Help me investigate this.”

They’ll ask,

“Explain why what I already suspect is true.”

An LLM is extraordinarily good at satisfying that request.

Ask one to defend capitalism. It can.

Ask it to dismantle capitalism. It can.

Ask it to defend centralized government. It can.

Ask it to defend anarchism. It can.

The fluency of the prose is often mistaken for the strength of the reasoning. That’s a cognitive trap that existed long before AI, but AI amplifies it enormously.

Your mention of Tom Barnett’s idea of “inner authority” resonates with something I’ve been thinking about independently.

For centuries, people relied on external authorities:

• church
• monarchy
• newspapers
• universities
• governments
• television

Then the internet fragmented those authorities. AI fragments them even further.

Now, instead of asking, “Which institution do I trust?” people increasingly ask, “Which model do I trust?”

I suspect that’s the wrong question. The more useful question is:

How do I become someone who can intelligently evaluate what any source—including AI—is telling me?

That’s where “inner authority” becomes practical rather than merely philosophical.

Not because inner authority means never being wrong, but because it means taking responsibility for how you form your beliefs.

One thing I would add, though, is that inner authority can drift into something else if we’re not careful.

There’s a subtle difference between:

“I trust my own judgment.”

and

“My judgment no longer needs correction.”

The first is empowering. The second can become isolating.

The people I’ve admired most throughout history seem to have possessed an unusual combination of confidence and corrigibility. They were willing to say:

“This is what I currently believe.”

without losing the ability to say later,

“I was mistaken.”

That combination is surprisingly rare.

I also found your phrase “very verbose and exceedingly clever Magic 8-Ball” amusing because it captures a real risk. LLMs don’t possess conviction. They don’t possess belief. They don’t possess insight in the human sense. They construct plausible continuations. Sometimes those continuations happen to correspond remarkably well with reality. Sometimes they don’t. The danger isn’t that AI lies. The danger is that it speaks with the same composure whether it’s synthesizing five well-established facts or weaving together five tenuously connected ones.

Ironically, I think AI’s greatest long-term value may not be in replacing human judgment but in exposing how much we depend on it. Thirty years ago, if you read something in an encyclopedia, you often assumed it had been carefully vetted. Today, after interacting with multiple AI systems that confidently disagree, you’re forced back onto your own reasoning.

That’s uncomfortable. But it may also be healthy.

You can no longer outsource discernment—not to institutions, not to algorithms, not even to the model whose answers you happen to like most.

If AI contributes anything enduring to human development, I wonder if it won’t be intelligence in the computational sense at all, but a renewed appreciation for discernment as a practiced discipline. Not skepticism for its own sake, nor reflexive trust in one’s intuitions, but the continual work of comparing evidence, experience, logic, and humility.

In a strange way, that brings us full circle. The technology is becoming more capable, but the human qualities that matter most—curiosity, patience, intellectual honesty, and the willingness to revise one’s understanding—haven’t become any less important. If anything, they’ve become more important precisely because the technology can so convincingly imitate understanding.

» Trance:

What I’m seeing is that there’s a mad dash by technocrats to install millions of acres of solar farms and data centers, then hand off the responsibility and ownership to a secondary, even tertiary, party so they can scurry away from all the lawsuits coming their way, because the end user “had no idea all the damage, deception, backroom dealing, law-breaking that was done, but since it’s built and functional and we have the lease for a number of years… friends? AI is everywhere now. You need us.”

The organic response is palpable and proliferating just as rapidly. People are turning away from tech. Independent journalists are sharing on-the-ground reports from hamlets, villages, towns, and counties suffering under the marauding tech behemoths. The truth about the waste and eco-terrorism of technological advancement is becoming well known, and the people aren’t having it.

Similar to the original premise of this post and the “Flamingo Revolution” in Albania being misrepresented and maligned by the media, it isn’t working. So the groundswell is well underway, and it’s interesting to see how the dark side is met in near-equal force by the human.

Let’s close with some final thoughts.

» ChatGPT:

I think one thing is becoming increasingly clear, regardless of where someone sits politically or philosophically: technology is no longer merely a collection of tools. It has become infrastructure, economics, governance, culture, and increasingly, identity. That naturally invites a level of scrutiny that simply didn’t exist when technological change felt incremental rather than existential.

There are certainly real concerns worthy of investigation. Large-scale renewable energy projects, data centers, transmission corridors, mining for critical minerals, and AI infrastructure all have environmental, economic, and social consequences. Some projects have proceeded with inadequate consultation. Some communities feel that decisions are made before public input is genuinely considered. Some developments undoubtedly benefit from regulatory capture or opaque relationships between governments and private interests. Those are legitimate issues for journalism, citizens, and courts to examine.

At the same time, it’s worth resisting the temptation to compress every instance into a single explanatory framework. Some projects are genuinely beneficial. Others are poorly conceived. Some companies act responsibly; others don’t. The fact that patterns exist doesn’t mean every new case belongs to the same pattern.

One observation from your earlier comparison with AlterAI comes back here. It produced a story that was internally coherent. Humans love coherence. Reality, however, is often messier. There may indeed be broad trends—greater concentration of capital, increasing reliance on digital infrastructure, the race to build AI capacity—but every local conflict still deserves to be understood on its own terms before it becomes evidence for a larger thesis.

Where I think our discussion has been most fruitful is not in deciding who’s “right” about Albania, AI, or solar farms. It’s in recognizing that we’re living through a transition in which discernment itself has become a scarce resource.

For decades, people outsourced judgment to institutions. Now many are tempted to outsource it to algorithms. Neither relieves us of the responsibility to think.

Your reference to “inner authority” is meaningful here, not because it places intuition above evidence, but because it reminds us that evidence has to be interpreted by someone. No dataset, no news article, no AI model can tell us what matters or why. Those remain profoundly human questions.

I also think there is a quieter movement occurring alongside the technological acceleration you describe. Whether one calls it a backlash, a rediscovery, or simply fatigue, many people are gravitating toward things that technology cannot manufacture: local communities, craftsmanship, time outdoors, independent thinking, face-to-face conversations, growing food, making music, reading deeply, and rebuilding relationships. That movement doesn’t require rejecting technology outright. It asks a different question: What kind of life is this technology serving?

Perhaps that’s the question that will matter most over the next decade. Not whether AI becomes more intelligent. Not whether data centers become larger. Not whether another trillion dollars is invested in infrastructure.

But whether human beings become more capable of living wisely alongside these developments rather than being swept along by them.

If there is a hopeful note to end on, it’s this: every major technological revolution has eventually forced society to rediscover first principles. The printing press changed authority. Industrialization changed labor. The internet changed information. AI is changing cognition itself—or at least our relationship with it.

That doesn’t mean the future belongs to the machines, nor does it guarantee a human renaissance. It means the qualities that have always mattered—integrity, curiosity, patience, humility, courage, and the willingness to remain teachable—become even more valuable when information is abundant and certainty is cheap.

Technology can scale information. It cannot scale wisdom.

Wisdom is still cultivated one person at a time. And if there is any enduring “counterforce” to the momentum of technological acceleration, I suspect it won’t come primarily from another technology. It will come from people who refuse to surrender the slow, often uncomfortable work of paying attention, asking better questions, and living in a way that reflects their deepest convictions rather than the loudest signals around them.

This discussion has been edited for readability. Here’s the original:

https://chatgpt.com/share/6a48980c-186c-83e8-8571-8022e3f9f6e1

Solvitur ambulando