S16 E38: Bitcoin Payments in 2025 with Sergej Kotliar & Matt Ahlborg
The audio keeps playing while you browse the site.
Read the transcript16,485 words, every line links to the minute it was said
Ask this episode
"I expected Lightning to kind of, you know, within a year or two, to be 50 and then 80 and then 95% of all transactions with internet money." Sergej Kotliar, CEO of Bitrefill, said that to me about a forecast he made years ago, and the figure he gave for how things turned out is about 5% of gross payment volume on his own platform.
Sergej was on the show for the third time. Next to him this time was Matt Ahlborg, a former Bitrefill data employee who now runs PPQ.AI, a pay-per-query AI service that takes bitcoin, and who was on for the first time. I wanted them together because between them they can see what people pay with when nobody is making a speech about it: Sergej from one of the oldest places to spend crypto on real things, Matt from a younger service that sells chatbot and image-generation credits. We ran just under two hours. Sergej's battery dropped to 4% around the 1:32 mark and he signed off at 1:40, so the last stretch is Matt and me talking about AI and privacy while the host fights the clock.
From 95% Lightning to "Many chains, many coins world"
I opened with payment statistics, since Bitrefill is about the only company that will give a straight breakdown. The ones we pieced together: Bitcoin's share of Bitrefill volume is 25 to 30%, by Sergej's estimate. Matt put the Ethereum ecosystem, stablecoins included, at 30 to 35%. Solana is about 10% (both of them agreed on that), Tron USDT 7 to 8%, and the rest is legacy coins, mostly Litecoin, with Dogecoin and Dash trailing it.
Sergej's summary of the shift: "we've, in the past years, moved from a Bitcoin maximalistic world to a post-Bitcoin maximalistic world." He described the share curve as "it used to be 90, then 80, then 60, then 50% Bitcoin," and called the destination a "Many chains, many coins world." On Lightning he was brief about the disappointment: "No mega rocket ships."
I asked whether the data could be trusted at that level of precision, and Sergej declined to oversell it: "that the data is not currently trustworthy, unfortunately." Later he added that "detailing with more precision always risks introducing more fake-news-ness than necessary." Matt's contribution was a reason the numbers blur at the edges: "Dashboard usage is quite high, where users are spending from their top-up balance," so the coin someone paid with and the coin they originally funded the account with can be different things.
Matt also had a practical complaint about Tron. "Tron is very expensive to use," he said, unless you stake Tron to get around the fees. Sergej's version of the same observation was about culture: "It's not like there are Tron maximalists or whatever." And the bigger ecosystems, in his words: "Like today's many crypto ecosystems, like Solana, like, they don't think about Bitcoin at all. Like, that's their entire thing."
Full RBF and the Zero-Conf Double Spend
The place where Sergej and I have history is full RBF. We covered it in S14 E4: Peter Todd vs Sergej Kotliar on Full RBF, and this time I told him up front that back then I was "naive and sort of sided with Peter."
Sergej's account of what happened after the policy change: before it, Bitrefill accepted zero-conf Bitcoin payments, with "years would go, millions of transactions without any double spends." The flow ran "without any incident," and afterward, in his words, "It lost money exactly once." That was a zero-conf double spend, the thing full RBF made easy to try. His verdict on the change itself: "From my perspective, Bitcoin became worse."
He also tied it to small payments in general. "The small transactions haven't really come back this time," he said, and Matt backed that up from the user side: "if they can't use the main chain, then they're just going to stop using Bitcoin or crypto full stop for the particular use case that they were doing." Their point, as I understood it, is that someone who cannot pay a small invoice on the base layer does not necessarily migrate to Lightning, they just go elsewhere.
I corrected Sergej on some of the history of the debate itself, and on the story of who said what when. His position on Peter, delivered without heat: "He said many things that were all right," and "which also has a certain element of truth." I raised the old Chris DeRose thesis that the morality of Bitcoin is whatever makes the number go up. Sergej's reading of the same thesis is where he calls Bitcoin "communism," a label I will let sit as his rather than mine.
"Decentralization is expensive, always, by definition"
I asked whether he believes in decentralization anymore. Sergej answered at 29:24: "No, of course not," followed immediately by the argument behind it, "Decentralization is expensive, always, by definition," and "the centralized option can always do everything that the decentralized option can do, but not vice versa."
His read of the competition is just as flat: "Bitcoin can't compete with that," and "it's hard for Bitcoin to win, maybe." Matt came at it from the opposite direction. He had always assumed regulators would come down harder on centralized versions of Bitcoin: "I always pictured that the more centralized versions of Bitcoin would have been cracked down upon harder than they have been up until now."
Stablecoins are where Sergej thinks this has already played out. "Stable coins in the current form are fully trusted," he said, and then gave Tether a job description: "the store of value, the unit of account and the medium of exchange of the crypto ecosystem." Bitcoin, on his telling, gets a different job, "the blue chip investment coin." He expects a future where "we're all going to go underground and become maximalist and privacy-focusing" if the pressure continues.
That is when I brought out the advice that I hear in the Bitcoin crowd, "never spend your Bitcoin," and pushed back on it. Sergej's description of the logic: "You should never spend your Bitcoin because that just drives price up and somebody else will spend the Bitcoin instead." Matt's view from the customer side: "It's the price action, generally speaking, when the price goes up for whatever crypto asset we have more customers coming to spend." And the growth, he said, is "slowly and steadily."
I also asked which users actually move coins around. Sergej's answer was NFT people: "Because you know, which group uses wallets a lot? NFT people. They have all their coins in a wallet. And they swap and trade and send to each other." That led into the Adam Back tweets I remember, and my line that "all it took was a couple of tweets from Adam Back saying, NFTs are scams." Matt's stance on all this is that the memetic energy comes from places Bitcoiners dislike: "The thing that really builds network effects for a lot of these tokens is the degen stuff. That's the catalyst that really brings it."
A Chain Is a Maintenance Contract
I asked Sergej the question I ask every payments company: why no Bitcoin Cash, and no Monero? His answer was about work rather than ideology. "Supporting a chain means maintenance duty, it means a whole new category of security considerations." He did not rule anything out forever: "we should never say never."
Matt, who has the numbers for his own side, was more direct about where he would put engineering time: "if I were in Bitrefill's shoes, I would add Bitcoin Cash way before I would add Liquid Bitcoin just in terms of number of users." On PPQ.AI, Monero is the second-largest payment method. "Monero and lightning were neck and neck," he said of the early days, and of its reputation outside crypto: "Monero is this buzzword in regulatory circles that says, Ooh, that thing's risky." For all that, on the Bitrefill side, "it's just extremely low usage."
Matt told me a small confession story about his own bag. He has about $1,000 in Liquid Bitcoin stuck in BTCPay Server because of how Blockstream's wallet handles it. Sergej had been poking at the Liquid mempool himself and noticed transactions of 40 to 47 satoshis carrying OP_RETURNs, which neither of us could fully explain. I later recorded Marc De Mesel in S17 E44 on the Bitcoin Cash bet and the privacy coin basket.
Sergej's pitch for what Bitrefill is: "we're the number one place in the world to buy things in the real world with your Bitcoin." The company is "closing in on 90 people." I asked him whether this makes Bitcoin stronger or just sits next to the old system, and he said "I don't know if we're making it stronger, but I would agree that we're not in that regard, really replacing things." He added a prediction I wrote down: "the libertarians are going to be upset with Bitcoin one day."
Bitcoin Culture, Seen From a Payments Desk
The conversation drifted from numbers to people. Matt described a "big cultural, like cognitive dissonance between like how Bitcoiners are supposed to think" and how they act. Sergej's version: "there's a lot of get-rich-quick mentality in the Bitcoin space." He also thought the industry's own habits had hurt it, calling one episode "kind of an own goal." Sergej put the lasting pattern like this: "There's always been a friendship layer in the ecosystem and an in-crowd layer and so on."
I pressed Matt on whether some of this was plain dishonesty. He said "I don't think that a lot of it is dishonesty," and offered a personal data point about the whole exercise: "I hold significantly less dollars in my bank account than I would have in the parallel universe where Bitcoin didn't exist." He still thinks long-term: "Bitcoin is something that's going to happen over centuries." Sergej, for his part, thinks the meme itself carries the weight: "I think Bitcoin's meme is the best meme," while Matt added "there are hundreds of millions of people using crypto."
My own contribution, since I am apparently the character who keeps asking about morality: I brought up whether it is reasonable to change your mind on a technical issue like RBF once your own business is on the line, and whether the person doing it should admit that. I did not get a tidy answer and Sergej did not pretend to have one.
Private AI, Paid in Sats
When Sergej left, I turned to Matt and PPQ.AI. The pitch is that "you don't have to connect personally identifiable information such as a credit card or an email address." Pricing is per query, and "Our average user spending about four bucks a month." One feature he added because a customer asked: "he told me to add that like six months ago." Daily users are modest, "We have hundreds of daily active users now, not thousands," and he is happy to say so. There is also a small naming story: Sergej suggested "PPQQ" and Matt could not explain what the second "Q" would stand for.
On privacy he drew a line I appreciated. PPQ.AI does not store IP addresses, only country codes, and images are kept by the AI providers for up to seven days. About the claims themselves he said: "we're promising that we're not storing that on our servers, but we're not proving that we are not storing it on our servers." And about a competitor: "they over marketed the privacy aspect of their service." His view of the structure is that "we just seem like one super user from the perspective of the big AI company." He thinks a "direct competitor" is a real possibility, and that privacy coin support "doesn't start with just integrating privacy coins." Looking at the rest of the market, he said "the very few holdouts in the media generation space that still insist on a subscription" are the ones he is competing against on pricing.
I asked whether he would use an AI tool for anything sensitive. His reply: "no, not right now, not for me." Then I asked the question every host asks these days, whether I am out of work. Matt: "I don't foresee massive AI unemployment in the next five years, maybe in 10 years." He thinks "truly original ideas instead of just regurgitated crap" are where humans keep an edge. He is also not blind to the upside: "there are definitely moments where AI tools give you kind of the tingles on like this stuff is really magical in the way that it can teach you things really fast."
His own habit is to feed Google's NotebookLM about 50 research documents a day and have it produce 20-minute podcasts, which he listens to. He also told me about the startups that did not work, among them a Bitcoin remittance MVP and a personal finance calculator. I later recorded Milan De Reede in S17 E25 on Nano GPT and AI vibe coding.
The agent angle came from Sergej before he left: the first AI agent purchase on Bitrefill came through a ZBD wallet. He thinks the web is not ready for them: "are all kind of designed to keep robots out," and his closing sentence for the future was "a wallet in every pocket." Matt, on whether a chat box is how people will eventually buy things: "a little chat box the best way to buy gift cards? Probably not."
Where we ended up disagreeing is the one thing neither side moved on. Sergej holds that Bitcoin is a blue-chip investment rather than a payment network, that decentralization is expensive, and that "You should never spend your Bitcoin" is a defensible default for holders. I still think a currency that nobody spends has failed at the job it was invented for, and that Lightning's 5% of his volume is a problem the Bitcoin side should be embarrassed about rather than a market verdict to shrug at. He thinks I am being sentimental, and I think he has made a good argument for why a company can thrive while the thing it accepts does not.
Listen to Sergej Kotliar & Matt Ahlborg on Spotify, Apple Podcasts, YouTube, and More
Time stamps:
- 00:02:20– Bitrefill’s Lightning Network Experience & Payment Trends
- 00:04:13– Shift from Bitcoin Maximalism to Multi-Chain Usage
- 00:05:47– Tron, Ethereum, and Payment Method Comparisons
- 00:10:04– Defining Post-Maximalism
- 00:11:08– Changing Morality and Ideology in Bitcoin
- 00:13:12– Proof of Reserves & Bitcoin Company Transparency
- 00:14:15– Bitrefill’s Payment Method Rankings
- 00:18:00– Payment Method Percentages & Dashboard Usage
- 00:20:21– Decline in On-Chain Bitcoin Payments
- 00:24:46– Popular Wallets for Bitrefill Purchases
- 00:27:53– Bitrefill’s Business Growth & Marketing
- 00:29:04– Adoption Pace & Decentralization Tradeoffs
- 00:31:35– Centralized vs. Decentralized Chains & Security
- 00:33:51– Stablecoins’ Role in Crypto Payments
- 00:35:25– Bitcoin as Store of Value vs. Stablecoins as Medium of Exchange
- 00:36:26– American Influence & “Never Spend Your Bitcoin” Narrative
- 00:38:47– Bitcoin Spending Trends & Storage Habits
- 00:43:34– Crypto Usage Beyond Bitcoin & Banking Systems
- 00:44:24– Bitrefill’s Current Offerings & eSIM Growth
- 00:47:50– Integration with Traditional Payment Networks
- 00:49:29– Loss of libertarian ethos and Bitcoin community fragmentation.
- 00:51:41– Support for Privacy Coins & Chain Selection
- 00:57:36– Monero Usage on PPQ.AI & Regulatory Perceptions
- 01:00:01– Liquid Bitcoin & Its Limited Adoption
- 01:05:16– NFTs, Gambling, and Network Effects in Crypto
- 01:07:14– Changing Narratives & Post-Rationalization
- 01:09:34– Bitcoin’s Utopian Vision & Real-World Challenges
- 01:13:06– Decentralization’s Gradual Growth & VPN Analogy
- 01:17:18– PPQ AI: Naming, Features, and Privacy
- 01:22:03– AI Model Variety & User Experience on PPQ AI
- 01:25:02– Image & Video Generation with AI
- 01:27:22– AI’s Impact on Employment & Content Creation
- 01:29:42– AI, Automation, and Crypto Payments for Robots
- 01:34:02– How PPQ.AI Handles User Data & Privacy
- 01:37:22– Memory Features & User Profiling in AI
- 01:39:19– Bitrefill’s Vision for the Future
- 01:40:29– PPQ.AI’s Business Model & Growth
- 01:44:18– PPQ.AI Model Access, Subscriptions, and Competitors
- 01:52:03– PPQ.AI’s Elevator Pitch & Referral Program
Verify this episode
This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.
- SHA-256
c32eaa1a0828917620040266470ec3ededee6742940aa5f024e9d407b1911e86- Proof
- 2025-08-12_s16_e38_sergej_kotliar_matt_ahlborg_on_bitrefill_payment_stats_ppq_ai_final_v2.mp3.ots · all episode hashes (proof)
- Status
- Anchored in Bitcoin block 974,312
- Timestamped
- After release. The episode came out on 12 Aug 2025, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
- Download the MP3 and the proof into the same folder.
- Run
sha256sum 2025-08-12_s16_e38_sergej_kotliar_matt_ahlborg_on_bitrefill_payment_stats_ppq_ai_final_v2.mp3(on a Mac:shasum -a 256 2025-08-12_s16_e38_sergej_kotliar_matt_ahlborg_on_bitrefill_payment_stats_ppq_ai_final_v2.mp3). The result must match the SHA-256 above. - Install the OpenTimestamps client (
pip install opentimestamps-client) and runots verify 2025-08-12_s16_e38_sergej_kotliar_matt_ahlborg_on_bitrefill_payment_stats_ppq_ai_final_v2.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node,ots info 2025-08-12_s16_e38_sergej_kotliar_matt_ahlborg_on_bitrefill_payment_stats_ppq_ai_final_v2.mp3.otsshows the block height to look up in any block explorer.
Related
S14 E4: Peter Todd vs Sergej Kotliar on Full RBF
In the first half of 2023, before ordinal inscriptions emerged and Drivechains stole the spotlight for the rest of the year, Full…
S17 E29: Grafton Clark on Vibe Coding, Privacy & Bitcoin Culture
Some podcast episodes are about a product, a protocol, or a company. This one is about a feeling: the feeling that the…
S17 E18: Dr. K on Scaling Proof of Work, Soaping the Competition with Quai & Unifying Physics
Dr. K (Karl Kreder), architect of the Quai Network, came back for his second appearance on Bitcoin Takeover to defend his…
Comments
Comments run on my own server. No trackers, no tracking cookies.