What if every click you made online cost just a fraction of a penny? What if your favorite news site, your go-to streaming service, or even your daily email usage could be paid for at tiny increments, rather than one big chunk at the end of the month? This vision—where nearly every digital interaction could be monetized by “micropayments”—has hovered over the internet economy since its earliest days. But as Nick Szabo’s seminal 1999 paper, Micropayments and Mental Transaction Costs, pointed out, there’s a lot more than technology standing in the way.

Twenty-five years on, Szabo’s warnings about mental transaction costs—the cognitive overhead of deciding whether something is worth paying for—still resonate. Even as developments like AI-based “intelligent agents” and Bitcoin solutions such as the Lightning Network promise frictionless micropayments, Szabo’s observations remain crucial to understanding why this idea hasn’t fully taken flight, and whether that might finally change.

Below, we’ll examine:

• The core arguments from Szabo’s 1999 paper

• Why micropayments remained on the fringes for decades

• How AI and Bitcoin’s Lightning Network attempt to overcome these barriers

• Whether mental transaction costs can, at long last, be reduced enough to make micropayments mainstream

The Paper That Defined the Dilemma

In Micropayments and Mental Transaction Costs, Nick Szabo pinpointed a truth that technologists often overlooked: while computational costs (like processing payments, preventing fraud, or validating cryptography) can be driven down, the mental overhead of deciding, monitoring, or worrying about every tiny expense remains stubbornly high.

“Customer mental transaction costs will soon dominate the technological transaction costs of the payment system used in the transaction (if they don’t already), and micropayment technology efforts which stress technological savings over cognitive savings will become irrelevant. ”

- Nick Szabo, Micropayments and Mental Transaction Costs (1999)

Szabo’s core argument is that for most consumers, there’s a cognitive “hassle factor” in even the smallest payment decisions. Asking yourself, “Is this article worth 2 cents? 5 cents? 10?” quickly leads to fatigue, overshadowing the supposed simplicity of micropayments. Instead, consumers gravitate toward flat fees and all-you-can-eat bundles, even if those end up costing slightly more in the long run. The mental relief of knowing that you won’t be nickel-and-dimed with every click is simply more valuable than the few pennies saved.

Sources of These Cognitive Costs”?

3 points are listed in the paper, but they can be many more.

1. Uncertain Cash Flows

Consumers rarely have perfect foresight into exactly how much they will earn or spend at any given time. Flat fees or bundling reduce the stress of planning and budgeting for these uncertainties.

2. Assessing Product Quality

In many online purchases—especially digital goods—you can’t know the true “quality” of what you’re buying until you’ve used it. Whether it’s an article, a game, or a movie, the mental effort needed to decide “Is this worth x?” every time you click can be more expensive than the micropayment itself.

3. Decision-Making Complexity

Our brains are good at making quick calls when stakes are high or options are few, but terrible when we have infinite micro-decisions.

Why Micropayments Stalled—Despite New Tech

1. The Early “Internet Payment” Hype

In the late 1990s and early 2000s, the internet was hailed as a new frontier for micro-billing. Systems like NetBill, Millicent, and PayWord promised frictionless flows of tiny sums. The dream? Artists, newspapers, and website owners would all be paid directly for each page view or each minute of content consumed.

But even as processing costs and fraud got more manageable, user adoption never reached critical mass. Szabo’s mental transaction cost argument largely explains this: Consumers found it simpler to deal with one monthly subscription than to handle countless pennies flying out of their digital wallets.

2. The Rise of “Free” Services Funded by Ads

Search engines, social media, and news sites gradually adopted a free-to-consume, ads-supported model. Why? It’s easy on the consumer’s mind—no sign-up or micro-accounting for every page load. Meanwhile, the site owner monetizes your attention via advertisements.

Even premium content gravitated toward low-friction paywalls and subscription models. Once the mental load of frequent, tiny payments was replaced by a single monthly charge, customers complained less and paid more consistently.

3. “Intelligent Agents” and AI: Early Promises, Slow Results

Szabo also anticipated solutions like “intelligent agents” that could, in theory, handle many micro-decisions on behalf of the consumer. The idea was that an AI could internalize your preferences (“I like reading about finance, but only from reputable sources, and I’m willing to pay up to 10 cents an article.”) and then automatically approve or decline micro-charges.

Yet building a truly personalized agent that doesn’t require continuous training and oversight—let alone potential conflicts of interest—has proven extremely challenging. For AI to manage micropayments accurately, it must grasp your tacit preferences and be trusted to act in your best interest.

Has Anything Changed in 25 Years?

While Szabo’s insights remain valid, the landscape in 2024 (and onward) does differ in a few important ways:

1. User Interfaces Have Improved

From intuitive mobile wallets to chatbots, user interface design is leagues ahead of where it was in 1999. Some friction has been removed: you can tap to pay, use passwordless logins, or integrate with wearables. But the cognitive overhead—the act of deciding whether a purchase is worthwhile—hasn’t vanished. Even a single tap is too much if you have to do it hundreds of times a day.

2. Blockchain & Cryptocurrencies

The Lightning Network has aimed to fix payments by enabling near-instant transactions with very low fees. It doesn’t solve the core argument of the paper, which assumes technical transaction costs are zero. But the Lightning Network is the current best standard and protocol on the internet for open, interoperable money to flow on the internet.

3. AI Enters The Chat

Tools like ChatGPT, advanced personalized recommendation engines, and agent frameworks have made it possible to tailor experiences more deeply to each user. In theory, an AI assistant could learn your tastes or budgets so well that you’re rarely disturbed with micro-approval prompts, or can automate them entirely within a certain budget. However, building up that trust in an AI agent remains a hurdle. The question moves from “Is this worth it?” to “What is my AI agent doing?”.

Looking Ahead: Are We Ready for a Micropayment Renaissance?

For mass adoption to happen, people need to avoid feeling nickel-and-dimed at every turn. Even if the technical fees are near zero, the mental transaction cost can make micropayments feel cumbersome. Making micropayments as invisible as possible, while keeping track of the value being exchanged, is therefore crucial.

Getting micropayments right will likely require a rethinking of business models, there are exciting examples where micropayments are emerging as a viable strategy:

• Pay-Per-API Call

In the AI SaaS world—micropayments are already thriving (called credits or tokens). Because companies evaluate usage strictly on ROI and business needs, they’re less deterred by the mental friction that keeps consumers at bay. They use just as much as they need in real-time.

• Tips & Donations

Small, voluntary payments for creators or open-source projects can work precisely because they don’t trigger the same sense of obligation. Users donate out of gratitude or community spirit, making micropayments feel more like a gesture than a forced charge. Stacker News and Nostr have been pushing this paradigm forward leveraging the Lighting Network.

Clever Design for Seamless Experiences

No matter the business model, user experience design is key to making micropayments practical. The simpler the interface, the more “invisible” the payments become. Some ideas include:

• Automated Rules & AI: Let users set broad preferences (“I don’t mind spending up to $2/day on premium articles”) and rely on an intelligent agent to handle decisions in the background.

• Bundled Invoices: Aggregate multiple micro-charges into one easy-to-understand statement, reducing the mental toll of each individual transaction. Ideally, this would be a standard and cross-product, instead of itemized in one niche or vertical.

• Intuitive Feedback: Offer clear yet minimal prompts—like a progress bar of monthly spend—that helps users track costs without being overwhelmed.

Overcoming the cognitive barriers identified by Nick Szabo demands not only faster, cheaper transaction rails but also thoughtful design that caters to real human psychology. When these elements come together—AI-based automation, usage-based models that don’t feel invasive, and a user interface that’s nearly frictionless—micropayments could see a genuine renaissance.

Conclusion: Szabo’s Insights Still Rule

Nick Szabo’s 1999 paper has proven remarkably prescient and held up after all these years. Even as technology has advanced—faster internet speeds, blockchain-based payment rails, and sophisticated AI—the central problem remains:

People don’t want to think about small payments all the time.

It’s not just about software or cryptography; it’s about the psychology of how we value attention, convenience, and certainty. Micropayments can succeed only if these mental costs can be minimized or “bundled away.” AI agents and the Bitcoin Lightning Network are crucial new pieces of the puzzle, but their success hinges on delivering a user experience that hides or automates micropayment decisions altogether.

Will the next 25 years finally bring an era where micropayments flourish? Possibly—if we figure out how to make paying a fraction of a penny feel as effortless as a monthly subscription. Even then, we might realize that micropayments simply become one more arrow in the quiver of payment models, coexisting with ad-based, subscription-based, and outright “free” offerings.

But for now, Szabo’s warning stands: a world of pure micropayments still collides with human psychology. Our mental transaction costs are real, and if the solutions of the future—be they AI, Lightning, or something else entirely—don’t address our deeper preference for simplicity, micropayments will remain an intriguing idea that never quite becomes the default.

References & Further Reading

• Szabo, N. (1999) “Micropayments and Mental Transaction Costs”

• Fishburn, P., Odlyzko, A. M., and Siders, R. C. (1997) “Fixed fee versus unit pricing for information goods”

• Nielsen, J. (1998) “The Case for Micropayments”

• Rivest, R. L. and Shamir, A. (1996) “PayWord and MicroMint—Two Simple Micropayment Schemes”

This is a guest post by Jacob Brown. Opinions expressed are entirely their own and do not necessarily reflect those of BTC Inc or Bitcoin Magazine.

Bitcoin’s decentralized consensus mechanism works based on some cleverly crafted incentive structures. The first and fundamental rule is that the chain with the most work is the correct one. This single rule obviates the need for a central arbitrator, determining which chain is correct as a function of the efforts of thousands of decentralized parties, each trying to extend the blockchain. The subsidy to miners keeps moving the blockchain forward, creating painful opportunity costs for miners who don’t mine the tip. These mechanisms, together with the difficulty adjustment, set the game theoretical framework for a chain that has marched forward, 1 block at time, with near 100% clarity for the last 15 years.

The only caveat is that if one miner or coalition of miners is able to marshal more than 50% of the hashrate, they will have the ability to overwrite recent blocks, prohibit other miners from writing future blocks, and determine which transactions are recorded in the canonical ledger. This would be a disaster, obviously; the entire point was to avoid a situation in which a single party was in control. So the ultimate binding piece of the game theory designed by Satoshi is that there is some incentive to prevent this from happening. As described in the whitepaper:

The incentive may help encourage nodes to stay honest. If a greedy attacker is able to assemble more CPU power than all the honest nodes, he would have to choose between using it to defraud people by stealing back his payments, or using it to generate new coins. He ought to find it more profitable to play by the rules, such rules that favour him with more new coins than everyone else combined, than to undermine the system and the validity of his own wealth.

He ought to find it more profitable to play by the rules

Indeed, this is the bedrock for all of the game theory in Bitcoin. Bitcoin makes sense if and only if, at any point in time, at least 50% of the miners are incentivized to stay honest. This has been the case since 2009.

An underdiscussed, but perhaps most crucial part of the theory is the reason why he ought to find it more profitable to play by the rules. The answer, in 2009, 2010, 2011, and every year since has always been the same: Because if he didn’t, it would break. If it breaks, the Bitcoin experiment is over and the miner who did this would become the proud owner of a landfill full of worthless E-waste. This is what Satoshi was referring to, and this is why the community panicked in 2014 when the ghash pool exceeded 50% of the hashrate. The idea that one party (even if that is a pool) could take over the system represented such a disastrous failure mode that everyone tries to avoid it.

Built into the game theory is the understanding that theoretically someone could, perhaps with significant costs, direct over 50% of the hashrate to behave in a dishonest way, forcing a constitutional crisis. But the natural result of this crisis is mutual assured destruction for all miners and holders. This is the ultimate deterrent for misbehavior.

Note that the theoretical possibility of a 51% attack is eternally present, regardless of the current hashrate, costs of electricity, cooling or new ASICs. This is a tautological consequence of the fact that 51% < 100%: At any point in time, a pool could be created with malicious intentions, and 60% of miners could join this pool. The fact of the matter is that in recent times, 100% of the miners are electively mining the tip. It is always a matter of incentives, not physical plausibility.

For those outside the system, who own no ASICs, the security model prohibits them from attacking the system. But the security model is designed not only to protect from external threats (it’s an open system after all) it’s designed to protect from actors within the system as well. Miners don’t just protect the system from non-miners, they protect the system from other miners.

Consider selfish mining. This technique is mathematically demonstrated to give an advantage to a group of 34% of miners who execute this technique beyond a difficulty adjustment period. Selfish mining doesn’t involve explicit stealing or even censorship, just a better ROI for the miners who would form the coalition. Recent reports have put the miner share of the top publicly held mining corporations at close to 30% and growing. Toss in a few large private miners and we get to the selfish mining threshold. Does it seem like selfish mining is inevitable? All that is required is that a collection of miners comprising 34% to hop on a call and start the process; three weeks later they’re reaping the rewards. Yet so far no groups of miners have made an attempt to try this. Why is this?

Selfish mining would represent a major norm violation; crossing this line would lead Bitcoin into a nasty place where competing groups are slugging it out. The grand prize for the winner is monopoly control, under which the monopoly miner gets to keep all the fees and block subsidies, can ease down their hashrate to boost profits, and can even negotiate fees directly or even set their own fee rates. But this would be a disaster for Bitcoin; for this reason, nobody is initiating that call.

I wrote a chapter in my book about coalitional game theory, analyzing exactly this problem in regards to monopoly mining. The analysis boils down to a comparison of the profits accrued to a 51% coalition which splits the rewards from a monopolized chain, or the small profits accrued to the grand coalition if they stick to the competitive course. In the early days, the answer was clear: Monopoly mining would have destroyed everything, so there is no incentive for a coalition to form.

Enter USG

If the USG commits to a plan, over years and decades, to invest in Bitcoin, they will have created something which cannot fail. It simply cannot. Regardless of who mines Bitcoin, who is priced out, what parties use the chain, it cannot fail, and it won’t fail. If there is a constitutional crisis about mining, this crisis will be resolved and resolved in a very clear and definitive way.

There are quite a few ways to resolve a constitutional crisis, when you expand your window to include centralized options. In the early days these options would have been discarded as inferior to failure, but if failure is not an option, all options can and will be considered. A simple brute force assertion of 51% power by USG and US controlled miners is one option (this need not require censorial monopoly mining.) Another workable solution is a permissioned soft-fork which only allows new blocks by the publicly traded miners. Obviously, Proof of Stake is on the table. Another option would be to convert the UTXO set of Bitcoin into a CBDC whose transactions are confirmed by the Fed. This would bring Bitcoin to the masses at lightning speed and bring massive value to early holders.

The point is that under this regime, monopoly mining is no longer a failure per se. Any coalition of miners could pursue monopoly mining, starting with selfish mining and snowballing their coalition to 51%. As long as they don’t do anything that directly irritates the USG, they can’t break the system. If they achieve monopoly mining, the USG is still there, backstopping Bitcoin.

In short, the USG enmeshing itself with Bitcoin’s success decades into the future removes Bitcoin’s ultimate weapon against centralization; its option to fail.

It’s hard to imagine that miners who are fighting for tiny profit margins would continue with the decentralization theater, when they ought to find it more profitable to form a coalition and monopoly mine, which strictly speaking, isn’t even against the rules.

This is a guest post by Micah Warren. Opinions expressed are entirely their own and do not necessarily reflect those of BTC Inc or Bitcoin Magazine.

Bitcoin was forged to be unstoppable in a hostile environment, but let’s be perfectly clear: surviving and thriving are two different things. Just because Bitcoin can withstand severe political antagonism doesn’t mean we should want that antagonism, nor does it mean we shouldn’t do everything possible to foster a favorable environment that accelerates adoption. Believing otherwise is a misreading of the core ethos. The brilliance of Bitcoin is that it remains permissionless and decentralized no matter who fights it—but that doesn’t preclude us from working to ensure we have the most beneficial conditions for its long-term success.

In fact, public policy responses to regulatory and legislative inquiries have consistently reaffirmed these basics: Bitcoin’s strength is open-source software, self-custody, and a wide distribution of mining and node operators. In other words, it’s not about selling out. It’s about ensuring our governments understand the benefits of Bitcoin’s open design.

There’s a difference between “Bitcoin was built for a hostile environment” and “we should want a hostile environment.” Having an adversary-resistant architecture doesn’t demand that we sit back and ignore opportunities to reduce friction, whether in energy policy or everyday user experience. Yes, Bitcoin can and will survive if politicians and regulators turn hostile. But it’s short-sighted to treat hostility as a virtue.

Hostility might slow adoption, push development offshore, or scare away everyday users who aren’t ready for that level of conflict. Meanwhile, measured engagement with policymakers can prevent draconian bans, shape balanced regulation, and offer legitimate pathways for institutional capital to flow in—all of which can speed up global usage of Bitcoin. It’s not a betrayal of Satoshi’s vision to say, “We’d like Bitcoin to flourish under transparent, fair laws.” We want people to choose Bitcoin, not be forced into it by some catastrophic breakdown of the legacy system.

There is nothing “un-Bitcoin” about encouraging legislation that protects individuals’ rights to use and hold their own BTC, or that supports open-source development. We should be unapologetically active in these political arenas, because ignoring them won’t make them go away. It would only allow others—perhaps with very different agendas—to set the rules in ways that hamper privacy, hamper self-custody, or hamper innovation.

The key is remaining vigilant against compromises that undermine the protocol’s integrity. Building relationships with politicians or regulators doesn’t mean we’re begging for favorable carve-outs at the expense of censorship resistance. It simply means we’re making our voices heard. If we see demands for forcing protocol-level changes that are hostile to users, that’s where we must stand firm and say “No” for both practical and ideological reasons. But proactively sharing how Bitcoin mining can stabilize energy grids or how Lightning Network can provide near-instant payments is not a concession of Bitcoin’s ethos. It’s part of a rational strategy to help the public and policymakers understand the real value behind Bitcoin’s existence.

Misguided concerns about large mining operations kowtowing to regulatory pressure are not new. The reality is, Bitcoin’s design remains adversary-resistant: anyone can mine if they have the hardware and energy, and anyone can run a full node to enforce the rules, ensuring that no single miner can change the protocol. If some mining pools bend to censorship demands, other pools are attracted by fees to include those transactions. That’s exactly how Bitcoin is designed: routing around censorship with an anti-fragile, decentralized architecture.

Ironically, positive regulatory engagement can reduce centralization risks if it opens more states, countries, and smaller energy providers to hosting mining facilities. Diversity of geography and jurisdiction means no single entity or government can easily impose sweeping rules on the entire network. Again, “hostile environment survival” doesn’t mean turning away from pragmatic solutions that help decentralize hashrate.

It is true that privacy, scalability, and accessibility remain pressing challenges. This isn’t an either/or proposition: we can both engage with regulators to stave off ill-informed policy and focus on advancing privacy-preserving features and scaling solutions. The key is not to let the everyday politics overshadow the work that needs to be done on second-layer technologies like the Lightning Network or more user-friendly privacy solutions.

Developers are actively tackling these issues, from better cryptography to more intuitive Lightning wallets. We should be championing—publicly and politically—initiatives that keep self-custody at the forefront and keep third-party custodians optional. Spreading knowledge of “not your keys, not your coins” at the legislative level isn’t selling out; it’s ensuring that more people (including politicians) actually grasp the fundamental reasons Bitcoin matters.

It’s easy to look at the ecosystem—full of corporate players, lobbying efforts, and social media theatrics—and think it has lost its soul. But Bitcoin has always been full of diverse voices, many of which care about short-term profit. That was true in 2011, it was true during the block-size wars, and it’s true now. It hasn’t destroyed Bitcoin. The network’s fundamental robustness ensures that, if you want to hold your own keys and validate your own transactions, nobody can stop you.

The central promise of Bitcoin hasn’t evaporated, and participating in policy doesn’t have to mean capitulation. It’s simply another stage in Bitcoin’s evolution, one where we actively shape a better environment for the technology and the people who benefit from it. We should embrace that fight wholeheartedly, defend Bitcoin’s fundamentals, and keep building toward a future where censorship-resistant, peer-to-peer digital money is the global norm—not just a contingency plan for hostile conditions.

This is a guest post by Pierre Rochard. Opinions expressed are entirely their own and do not necessarily reflect those of BTC Inc or Bitcoin Magazine.