🧠 We Can’t Even Align With Ourselves, So How Will AI?

Sign Up | AI Toolkit | Meet The Team

Welcome back AI prodigies!

In today’s Sunday Special:

• 📜The Prelude

• 🪞Why We Can’t Align With Ourselves

• 🏠How Systems Misalign: The 2008 Financial Crisis

• 🤖Why Human Misalignment Undermines AI Alignment

• 🔑Key Takeaway

Read Time: 7 minutes

🎓Key Terms

• AI Alignment: Ensures AI Systems behave consistently with human intentions, values, and goals.

• Cognitive Dissonance: The mental discomfort that occurs when our beliefs are contradicted by new information.

🩺 PULSE CHECK

📜THE PRELUDE

We often tell ourselves we want to hit the gym, wash the dishes, or go to bed early. And yet, we find ourselves endlessly distracted: we delay, we binge, we procrastinate.

We may regret it, but we repeat it. This cycle raises a deeper, more unsettling question: “What do we actually want?” Our brain says one thing, while our behavior says another. This pattern makes at least one of the following statements true:

1. We aren’t always sure what we truly want in any given moment.

2. Our core values aren’t strong enough to resist immediate temptation.

If we can’t always act on our own values, how can we teach machines to do it, especially when we don’t always agree on what those values are? In other words, how can we expect machines to align with such fractured, unstable, or unclear values?

🪞WHY WE CAN’T ALIGN WITH OURSELVES

⦿ 1️⃣ The Evolutionary Trait of Short-Term Thinking.

The real challenge we face is this: we routinely fail to align our actions, choices, or behaviors with our stated values, beliefs, or goals. So, why do we do this?

For most of human history, life was short and filled with constant uncertainty. Back then, hunter-gatherers had an average lifespan of 33, nearly 40 years less than today’s average lifespan. Survival was precious: no shelter was certain, and no meal was guaranteed. During that time, delayed gratification was a risky proposition.

Today, success often depends on planning for the future and resisting the temptation of instant gratification. For example, the average retirement age in the U.S. is approximately 65, with about 60% of Americans funneling money into retirement accounts, such as a 401(k).

Despite these social shifts over time, our brains remain wired for immediacy, nudging us to prioritize short-term pleasure over long-term effort. This evolutionary trait helps explain why our behaviors often betray our stated intentions: we scroll social media feeds instead of studying or snack all day despite wanting to eat healthier.

⦿ 2️⃣ The Psychology of Cognitive Dissonance.

In 1957, American social psychologist Leon Festinger proposed a theory of Cognitive Dissonance: the psychological discomfort experienced when a person holds two or more conflicting beliefs. Crucially, he discovered that people typically don’t resolve this tension by changing their behavior to align with their beliefs. Instead, they unconsciously change their beliefs to justify their behavior.

For instance, we know that scrolling through social media feeds for hours can reduce our memory, attention span, and sleep quality. Still, we might rationalize the behavior: “I need to know what’s going on,” “Everyone does it,” or “It’s how I connect with friends.” Our beliefs mutate to preserve our behavior.

🏠HOW SYSTEMS MISALIGN: THE 2008 FINANCIAL CRISIS

⦿ 3️⃣ The Set Up.

Even when we act in alignment with our values, beliefs, or goals, things can still go terribly wrong. The 2008 Financial Crisis is a textbook example of how misaligned incentives can lead to catastrophic outcomes. What began with risky lending evolved into a global economic meltdown because each key player involved acted “rationally” within their own silo, without considering how their actions affected the entire System. To understand how Systemic Misalignment manifested, we must first look at the 5 key players involved and their incentives:

1. Mortgage Originators: They gave out Subprime Mortgages: high-interest-rate home loans given to risky borrowers with poor credit. In the years leading up to 2008, they didn’t care whether the risky borrowers could actually repay the loans because they immediately sold them to Investment Banks (IBs) on Wall Street. Their logic was: “I get paid when I issue the loan, not when it gets repaid.”

2. Banking Powerhouses: They bought thousands of high-interest-rate home loans made to risky borrowers and combined them into a new financial product called Mortgage-Backed Securities (MBS), which allowed institutional investors to bet on whether the mortgage payments within MBS would be paid off. Essentially, IBs made money by purchasing high-interest-rate home loans, repackaging them, and selling them to institutional investors. They relied on Credit Rating Agencies (CRAs) to give MBS high ratings. Their logic was: “As long as the financial product is rated safe, we can sell it at a premium to institutional investors and make a profit.”

3. Risk Rating Agencies: They were supposed to independently assess the risks of MBS. But they were paid by the very IBs whose MBS they were rating. To keep the IBs happy, they often gave overly optimistic ratings. Their logic was: “If we don’t rate these highly, the IBs will go to a competitor agency.”

4. Institutional Investors: These included Pension Funds, which manage retirement savings, and Insurance Companies, which must always have assets to pay out claims. Both are heavily regulated and required to invest in safe assets to meet their obligations. Because MBS received top ratings from CRAs, institutional investors bought them. Their logic was: “It’s rated highly, so it must be safe. Plus, regulations require us to hold high-rated assets like this.”

5. Regulators and Policymakers: Government regulators faced political pressure and societal expectations to promote homeownership in order to allow the housing market to continue booming. After all, buying a house is part of the American Dream. Their logic was: “Let the housing market work itself out and mature. It’s helping more people get homes and access credit!”

⦿ 4️⃣ Collapse and Resolution.

Eventually, the risky borrowers began to default on their mortgage payments. In other words, they stopped paying their monthly installments. Without mortgage payments from these risky borrowers, the value of MBS plummeted. To prevent total financial collapse, the U.S. government created the Troubled Asset Relief Program (TARP), a $700 billion bailout program to stabilize major banks and restore confidence in America’s financial frameworks.

The 2008 Financial Crisis revealed that when key players follow their own values, beliefs, or goals, their choices, actions, or behaviors can still destabilize the System. This illustrates a critical aspect of Systemic Misalignment: rational choices made in isolation can lead to catastrophic outcomes when combined. So, what’s this got to do with AI developments?

🤖WHY HUMAN MISALIGNMENT UNDERMINES AI ALIGNMENT

If we struggle to align with ourselves and fail to coordinate effectively with each other, how can we expect to align machines with human values?

At the individual level, human values are unstable, fragmented, and context-dependent. We rationalize past behavior to preserve comfort rather than face the truth.

At the group level, human misalignment becomes amplified. We rarely share a single, unified set of human values. And even when we do, underlying incentives like money, power, or survival push us to act against them.

The current landscape of AI developments mirrors this structure. Competing companies, institutions, and governments each pursue AI Alignment under their own definitions, incentives, and timelines relative to market pressures and profit motives. The result isn’t a unified AI Alignment effort, but rather many loosely coupled and potentially conflicting initiatives to guide AI developments.

This creates a Recursive Trap: misaligned humans train advanced AI models that reflect, mutate, and multiply their misalignment. The future of AI Alignment depends not just on better technical breakthroughs but on our collective willingness to confront human misalignment.

That means acknowledging that the very human values we want advanced AI models to reflect are often unstable, fragmented, and conflicting. In other words, what does it mean to align advanced AI models with human values that shift based on mood, attention, or social context? As long as human nature endures, AI Alignment will remain a moving target.

🔑KEY TAKEAWAY

We constantly struggle to align our own actions with our values. AI Alignment inherits this challenge: how can machines reflect human values if those values are unstable, conflicting, and distorted by incentives? Until we decode the forces that fuel human misalignment, aligning AI with human values may remain more myth than mission. In simpler terms, right now, it’s an ideal we’re chasing without truly understanding what it means.

📒FINAL NOTE

FEEDBACK

❤️TAIP Review of The Week

REFER & EARN

🎉Your Friends Learn, You Earn!

You currently have 0 referrals, only 1 away from receiving 🎓3 Simple Steps to Turn ChatGPT Into an Instant Expert.

Share your unique referral link: https://theaipulse.beehiiv.com/subscribe?ref=PLACEHOLDER