One balance, many positions Every liquidity provider runs into the same arithmetic. You have one balance and four ideas about where it should sit: a stablecoin pair that grinds out volume, a volatile pair with wide spreads, a tight range on an L2 where flow is picking up, a wider range as insurance. All four are reasonable. You get to fund one, or fund all four badly. That has been treated as physics. It is a design decision, and @1inch Aqua reverses it. The pre-funding assumption Outside crypto, nobody expects a quote to be pre-funded. A market maker quotes the same inventory on several venues at once - first fill wins, the rest of the quotes get pulled. A quote is an offer, not an escrow. DeFi inverted that. To advertise depth you had to hand the assets over first, because a pool needs custody in order to promise anything. Everything else follows: capital fragments across pairs and chains, each pool ends up thinner than it should be, and one balance produces exactly one position. The cost is not mainly a yield story. It is a depth story - the market sees less liquidity than actually exists. $542 million doing nothing In a @Dune resrarch commissioned for the launch, 85% of concentrated liquidity on major DEXs was used inefficiently in the first half of 2026 - about $1.6 billion of the $1.84 billion tracked. Roughly $542 million was, on average, entirely out of range for a full week at a time. The estimate for fees never earned is around $150 million a year. Half a billion dollars, in the right protocol on the right pair, unreachable for a week. Not mispriced, not exploited. Just committed somewhere it was not needed, because price moved and the balance could not be in two places. A registry, not a pool Aqua stops moving tokens. You connect a wallet, approve a balance, and create positions that are allowed to reference it. The contracts hold nothing - a swap can only pull what is actually in the wallet at the instant it fills, and revoking the approval stops new fills as soon as it confirms on-chain. Remove custody and the constraint that produced all the fragmentation goes with it. The same balance can stand behind several positions at once, across pairs and across 13 EVM chains, each deployment independent. 1inch's own example: a $100,000 balance backing three positions that together quote $300,000 of depth. Availability, not exposure Nothing is borrowed here. No credit, no interest, no debt position. Three positions quoting $100,000 each cannot pull $300,000 out of a wallet holding $100,000, for the same reason a market maker on five venues cannot sell the same inventory five times. Your worst case is your own balance. If a fill arrives and the balance behind it is already spent, the position pauses instead of liquidating. The failure mode is inaction, not loss. You choose the pair, range and fee, and close whenever you want. The security story is structural before it is procedural. The standard DeFi incident is a contract holding a very large balance; Aqua's hold none. And because every position has a single owner, there is no shared fee moment for JIT bots to snipe. Eight independent audits - Hexens, OpenZeppelin, Bailsec, Nethermind, Hashlock, MixBytes, Theori, Decurity - reduce risk. They do not eliminate it. The category question Products that matter rarely add a feature. They remove a requirement everyone had stopped noticing. Aqua removes pre-funding, and the consequence is an accounting one: quoted liquidity and locked capital stop being the same number. TVL as a proxy for depth, custody as the price of participation, range selection as a single bet on price - all of it assumed those two numbers were identical. There is a Merkl-powered incentive program running for early providers, with terms on the official @1inch blog. But the honest test is not reading about it. Open one position on a pair you already understand, keep your tokens where they are, and watch what one balance does when it is allowed to be in more than one place. Provide Liquidity:

One balance, many positions Every liquidity provider runs into the same arithmetic. You have one balance and four ideas about where it should sit: a stablecoin pair that grinds out volume, a volatile pair with wide spreads, a tight range on an L2 where flow is picking up, a wider range as insurance. All four are reasonable. You get to fund one, or fund all four badly. That has been treated as physics. It is a design decision, and @1inch Aqua reverses it. The pre-funding assumption Outside crypto, nobody expects a quote to be pre-funded. A market maker quotes the same inventory on several venues at once - first fill wins, the rest of the quotes get pulled. A quote is an offer, not an escrow. DeFi inverted that. To advertise depth you had to hand the assets over first, because a pool needs custody in order to promise anything. Everything else follows: capital fragments across pairs and chains, each pool ends up thinner than it should be, and one balance produces exactly one position. The cost is not mainly a yield story. It is a depth story - the market sees less liquidity than actually exists. $542 million doing nothing In a @Dune resrarch commissioned for the launch, 85% of concentrated liquidity on major DEXs was used inefficiently in the first half of 2026 - about $1.6 billion of the $1.84 billion tracked. Roughly $542 million was, on average, entirely out of range for a full week at a time. The estimate for fees never earned is around $150 million a year. Half a billion dollars, in the right protocol on the right pair, unreachable for a week. Not mispriced, not exploited. Just committed somewhere it was not needed, because price moved and the balance could not be in two places. A registry, not a pool Aqua stops moving tokens. You connect a wallet, approve a balance, and create positions that are allowed to reference it. The contracts hold nothing - a swap can only pull what is actually in the wallet at the instant it fills, and revoking the approval stops new fills as soon as it confirms on-chain. Remove custody and the constraint that produced all the fragmentation goes with it. The same balance can stand behind several positions at once, across pairs and across 13 EVM chains, each deployment independent. 1inch's own example: a $100,000 balance backing three positions that together quote $300,000 of depth. Availability, not exposure Nothing is borrowed here. No credit, no interest, no debt position. Three positions quoting $100,000 each cannot pull $300,000 out of a wallet holding $100,000, for the same reason a market maker on five venues cannot sell the same inventory five times. Your worst case is your own balance. If a fill arrives and the balance behind it is already spent, the position pauses instead of liquidating. The failure mode is inaction, not loss. You choose the pair, range and fee, and close whenever you want. The security story is structural before it is procedural. The standard DeFi incident is a contract holding a very large balance; Aqua's hold none. And because every position has a single owner, there is no shared fee moment for JIT bots to snipe. Eight independent audits - Hexens, OpenZeppelin, Bailsec, Nethermind, Hashlock, MixBytes, Theori, Decurity - reduce risk. They do not eliminate it. The category question Products that matter rarely add a feature. They remove a requirement everyone had stopped noticing. Aqua removes pre-funding, and the consequence is an accounting one: quoted liquidity and locked capital stop being the same number. TVL as a proxy for depth, custody as the price of participation, range selection as a single bet on price - all of it assumed those two numbers were identical. There is a Merkl-powered incentive program running for early providers, with terms on the official @1inch blog. But the honest test is not reading about it. Open one position on a pair you already understand, keep your tokens where they are, and watch what one balance does when it is allowed to be in more than one place. Provide Liquidity:

🚨🏧 BREAKING: Oobit turns on ATM cash withdrawals worldwide The Tether-backed payments platform @oobit now lets users pull physical cash by tapping their phone on an NFC ATM - no plastic card, no bank account. Funds come straight from the Oobit balance, with the card PIN unlocked by biometrics in the app. Limits: $250 per transaction, up to 3 withdrawals per card per day. Context: @oobit already covers 150M Visa merchants across 80+ countries. ATMs close the last gap - the exit into physical cash. In Oobit's core markets (LATAM, Africa, Southeast Asia) the ATM is still the main cash rail, and a bank account is far from universal. Platform activity over the past six months: +260%

Mastercard Q2 2026: what the network's report tells the private market @Mastercard quarterly report gets read as a stock idea. For a fund, that is the least useful way to read it. A public incumbent doing $9.3B in quarterly revenue is not a ticker, it is the best available dataset on which private companies someone is willing to buy, at what price, and in which layer of the stack margin actually exists. There are four such signals in this report, and none of them is about EPS. The category ceiling has moved On March 17, Mastercard agreed to acquire @BVNKFinance for $1.8B - $1.5B plus up to $300M in contingent consideration. It is the largest stablecoin infrastructure acquisition on record. The previous marker, @stripe $1.1B purchase of Bridge (@Stablecoin), held as the category ceiling for two years. For anyone holding a position in payment infrastructure, this repriced the exit math. BVNK processes over $30B in payments annually, which means the comp is a multiple of volume rather than revenue, and that is the anchor both founders and buyers will negotiate against in the next round. The ceiling rose 60% in two years, in a category where the existence of exits at all was in question not long ago. There are fewer buyers than it looks The second piece of news matters more than the first, and almost nobody picked it up. In May, Mastercard walked away from a minority investment in @zerohashx, a company it had earlier discussed acquiring outright for as much as $2B . @zerohashx went out to raise on its own instead, above a $1.5B valuation. Read that as an architecture decision. Mastercard chose a single integrated stack over a portfolio of bets on competing providers. The practical consequence for a fund: this category just lost a strategic buyer, and that buyer has already spent its budget. The first company in the segment exits at $1.8B, the second raises, and the third and fourth compete for an acquirer who is no longer in the market. Exit concentration risk here is higher than the size of the market suggests. The margin is not in the rail Mastercard's payment network grows 8-12% currency-neutral. Value-added services - scoring, authentication, fraud prevention, data - grew 22% year over year in Q1, organically, and now account for roughly 40% of net revenue. So an incumbent running a 58% operating margin is showing you where its money is while simultaneously paying $1.8B not to build settlement itself. Both facts point the same direction. Settlement is being commoditized, and it is being bought. Software sold on top of the traffic is not. If your pipeline holds another cheaper, faster rail, that is a company acquired for its volume in the best case. A company selling risk data and compliance on top of someone else's rails gets acquired for its revenue. The authorization window is open, and closing Agent Pay is now enabled on essentially every Mastercard card globally, with Verifiable Intent layered on top as a tamper-resistant record of user authorization, plus a Crossmint partnership for blockchain execution. An agentic payment creates a problem of consent, not settlement: who authorized what, exactly, and how do you prove it in a dispute six months later. A stablecoin rail does not address that question at all. Mastercard is entering agentic commerce through authorization rather than through the transfer, and doing it with distribution across billions of cards. This is the most interesting open layer in payments right now and the one closing fastest. What to take from it Public markets are paying for the transformation: 25.8x forward earnings against 18.6x for the industry. That is the discount rate underneath the whole thesis - incumbents hold expensive paper and have every reason to buy infrastructure with it. So the useful question after this report is not whether to own $MA . It is which layer of the stack @Mastercard pays for next instead of building. This year the answer was settlement. Judging by Agent Pay, next time it will not be.

The fee was never the point, everyone assumes the story is cost. It isn't. The story is what disappears once cost stops mattering. For years @trondao identity was the small transfer - money sent home, sub - $1,000 payments that made up most of its volume. That is a real infrastructure with a use case that creates potential beyond remittances. What changed is not the price of a transaction. It is who has to think about one. GasFree lets a user move USDT without ever holding native token, sponsoring the network fee and taking $1.50 out of the transfer itself. On an average transfer of $16,300, that is 0.009%. Nobody has to learn what energy or bandwidth means. They just send. Remove that step and the customer changes. Nobody sends $16,000 to a relative. Those are payrolls, supplier invoices, treasury moves - flows that were always waiting for a rail where the fee was boring rather than cheap. The lesson travels past crypto. Infrastructure wins by becoming invisible, and the last thing to go invisible is almost always the price tag.

@TheBitcoinConf with @oobit, this has been possible for a long time You can pay with crypto (BTC, USDT, etc.) anywhere that accepts Visa - including Emirates tickets, and even get up to 10% cashback. so for me, this isn’t news, it’s just another step forward.

USDT's global supply barely moved in Q2 - roughly $184B, effectively unchanged. @trondao slice of it hit an all-time high $90B, about 47% of every USDT in existence. This is what happens when people are choosing a rail instead of inheriting one. The composition explains the choice. Around 93% of stablecoin transfer volume on @trondao is p2p, the highest of any chain. Among chains with native USDT issuance, TRON's share of sub - $1,000 transfers moved from 43% to 52% in a single quarter. Retail is not a segment here. It is the base load. Most networks treat that traffic as low value. TRON converted it into $89M in protocol fees, second only to Hyperliquid, and 34% of crypto card volume, the highest share of any chain. Card spend across all chains grew from $2.0B to $2.4B, so TRON took a bigger slice of a bigger pie. The lesson for anyone building payments: distribution beats architecture. Stablecoin flows settle where the users already are, and users stay where fees are predictable and sending $200 does not feel like a decision.

USDT volume on @Revolut rose to $855m per week amid news of the delisting. The monthly volume hit an all-time high of $2b An important point to note: this is not a surge in demand. It is an outflow. Users are withdrawing USDT to self-custody and other platforms while the window is still open. he peak will almost certainly occur at the end of July (when deposits close) and in the second half of August (the deadline for forced conversion), after which volumes at these addresses will drop to near zero.

Buying gold was never the problem. Moving it was. Turning ethereum:0x68749665ff8d2d112fa859aa293f07a622782f38 into cash meant selling it, converting it, waiting then withdrawing. Days for one transfer. @obchakevich_, in collaboration with Oobit, looked at how that became one move. → No exchange hop. No middlemen. No big cut in fees. → Money lands straight into local bank rails: PIX, ACH, SEPA... → You see what leaves and what arrives, before you confirm. XAUt now converts and settles directly into a bank account, in seconds. Full report below:

$USDT volume on Revolut hit an all-time high of ~$2B in July, peaking at $855m in a single week. Not a demand surge, pure outflow. Users are selling and withdrawing before the 31 Aug delisting deadline due to MiCA regulations in the EU. Credit to @obchakevich_ for the information and chart.

nobody tells you straight that most crypto exchanges and platforms will disappear. you just see new listings, the hype, and think this is forever. but the market works differently. over time only a few strong players will remain. regulations do their job. in the eu, mica from july 1 2026 put an end to hundreds of firms. no license - shut down. estimates say up to 80% of european crypto companies will not survive. binance restricted services in the eu, bitmart announced an orderly wind-down. these are not exceptions. this is a systemic weeding out of the weak. weak platforms lived on volumes, tokens and marketing. the strong built compliance, capital and reliable infrastructure. when the rules get strict, the first ones disappear. just like in traditional finance after every crisis: those who can afford to be fully regulated remain. this is not the end of crypto. this is its cleansing and maturing. the market becomes more mature. users get less risk and more trust. decentralized solutions can survive differently, but centralized exchanges, only the strongest ones. the future belongs to those who are not afraid of the rules, but use them as a filter. most will close. a few strong ones will win. and this is good for the entire industry. history always repeats itself.

USDT volume on @Revolut rose to $855m per week amid news of the delisting. The monthly volume hit an all-time high of $2b An important point to note: this is not a surge in demand. It is an outflow. Users are withdrawing USDT to self-custody and other platforms while the window is still open. he peak will almost certainly occur at the end of July (when deposits close) and in the second half of August (the deadline for forced conversion), after which volumes at these addresses will drop to near zero.

nobody tells you straight that most crypto exchanges and platforms will disappear. you just see new listings, the hype, and think this is forever. but the market works differently. over time only a few strong players will remain. regulations do their job. in the eu, mica from july 1 2026 put an end to hundreds of firms. no license - shut down. estimates say up to 80% of european crypto companies will not survive. binance restricted services in the eu, bitmart announced an orderly wind-down. these are not exceptions. this is a systemic weeding out of the weak. weak platforms lived on volumes, tokens and marketing. the strong built compliance, capital and reliable infrastructure. when the rules get strict, the first ones disappear. just like in traditional finance after every crisis: those who can afford to be fully regulated remain. this is not the end of crypto. this is its cleansing and maturing. the market becomes more mature. users get less risk and more trust. decentralized solutions can survive differently, but centralized exchanges, only the strongest ones. the future belongs to those who are not afraid of the rules, but use them as a filter. most will close. a few strong ones will win. and this is good for the entire industry. history always repeats itself.

🚨BREAKING: USDT volume on @Revolut has risen to nearly $2B in the first part of July amid news of the delisting. Based on my data, customers are actively exchanging and selling solana:Es9vMFrzaCERmJfrF4H2FYD4KCoNkY11McCe8BenwNYB on Revolut to beat the August 31 deadline, when access to USDT will be completely shut down on Revolut.

Our new report "Closing the Loop: The Self-Driving Landscape" is out now! AI can generate scientific hypotheses faster than laboratories can test them. Self-driving labs are designed to close that gap by automating the full experimental loop. The model chooses the next experiment, and connected laboratory equipment carries it out. The results feed back into the system and shape what it tests next. The stakes are especially high in drug discovery. Lead optimization alone can consume roughly three years. Bringing a drug to market takes 10–15 years, with average out-of-pocket and time costs of $2.6 billion per drug. Self-driving labs target the earlier experimental bottleneck. They could shrink individual cycles from months to days or hours. Running more experiments can also reduce the cost of each run by spreading the upfront cost of automation further. Every completed experiment adds to a structured record of what worked and what did not. That data improves the model’s next decision, creating a continuous learning loop between AI and the physical lab. AI has accelerated the generation of scientific hypotheses. Self-driving labs could accelerate the experiments that determine which ideas are worth pursuing.

If you look at numbers instead of hype, the picture is surprising: TRON consistently generates between $400k and $1m in revenue per day. Over the last 30 days - $25.9m in net revenue. The secret isn't technology. @trondao is the highway for USDT: the average daily transfer volume is $23B in the last 30 days. It's cheap and predictable. In emerging markets, it has long become the default payment infrastructure. People sending money to family across borders don't care about architecture - they care that the transfer arrives in seconds. Revenue is the most honest metric in crypto because it’s money users actually paid for the network to do its job. And by this metric, the L1/L2 market looks simple: there's TRON, and there's everyone else.

🚨BREAKING: USDT volume on @Revolut has risen to nearly $2B in the first part of July amid news of the delisting. Based on my data, customers are actively exchanging and selling solana:Es9vMFrzaCERmJfrF4H2FYD4KCoNkY11McCe8BenwNYB on Revolut to beat the August 31 deadline, when access to USDT will be completely shut down on Revolut.

If you look at numbers instead of hype, the picture is surprising: TRON consistently generates between $400k and $1m in revenue per day. Over the last 30 days - $25.9m in net revenue. The secret isn't technology. @trondao is the highway for USDT: the average daily transfer volume is $23B in the last 30 days. It's cheap and predictable. In emerging markets, it has long become the default payment infrastructure. People sending money to family across borders don't care about architecture - they care that the transfer arrives in seconds. Revenue is the most honest metric in crypto because it’s money users actually paid for the network to do its job. And by this metric, the L1/L2 market looks simple: there's TRON, and there's everyone else.