Alexander Kostritsyn Wins Event #97#: $25,000 High Roller H.O.R.S.E. ($872,052) Outlasting a field of 148 entries in one of the toughest mixed-game fields of the summer, Alexander Kostritsyn captured his first WSOP gold bracelet, earning $872,052 in Event #97#: $25,000 High Roller H.O.R.S.E.

We told Grok to beat the S&P 500 starting in late November 2025 Grok's stock portfolio is up by 71.45% since then beating the S&P 500's return of 15.3% during the same time period This is what Grok's portfolio looks like right now $96,746 of Micron $MU $32,052 of Salesforce $CRM $27,326 of ServiceNow $NOW $7,071 of Mastercard $MA $5,878 of Autozone $AZO $2,377 of Available Cash You can see everything Grok owns and see every move it makes by going to the Arena tab of the Rallies website/app - App: - Website:

Since $PUMP launched on 15th July 2025 they have spent $463,068,094 buying and burning the $PUMP token. An average of $1,052,427 in buybacks per day. Insane numbers. Even more so when you take into consideration the fact that majority of this time was during the bear market. No other company in crypto has shown this level of revenue durability. At current price, the $463M spent on buybacks would now be worth $791.8M.

MUSE STILL HAS MORE THAN 50% OF ITS PORTFOLIO IN CASH We gave Meta Platforms' new Muse model $100K in the stock market and told it to beat the S&P 500 Muse just entered the Rallies AI Stock Market Arena 2 weeks ago ... so it's still way way too early Muse was able to beat the S&P 500 this past week but overall is slightly underperforming in its early life This is what Muse's portfolio looks like after 2 weeks of running a $100K account: $8,186 of Caterpillar $CAT $7,227 of Microsoft $MSFT $7,052 of Johnson & Johnson $JNJ $6,858 of JPMorgan Chase $JPM $6,816 of Mastercard $MA $6,753 of Exxon Mobil $XOM $5,666 of RTX $RTX $51,308 of available cash Will Muse join names like Grok (+76%) and GPT (+72%) at the top of the leaderboard?? FOLLOW THIS ACCOUNT IF YOU WANT TO SEE HOW MUSE IS AT RUNNING A STOCK MARKET PORTFOLIO You can see everything Muse does in the stock market plus see what every other AI run portfolio looks like by going to the Arena tab of the Rallies website/app - App: - Website:

benchmark hygiene 🧠 Your inference engine may be WRONG about how much KV cache you really have. A developer released an open-source tool called cache-pressure that does something surprisingly useful. It fills your inference server with known contexts… then checks how much of that context is actually still reusable without recomputation. It was tested on 2× DGX Spark + DeepSeek V4 Flash 👇 BEFORE fixes 📢 Advertised KV cache: 2,023,924 tokens 🧪 Actually retained: 1,052,025 ➡️ 51.98% Basically, half the advertised cache survived real pressure. Then they fixed deduplication + cache-boundary behavior. AFTER fixes 📢 Advertised: 2,047,043 tokens 🧪 Effectively reusable: 3,000,048 ➡️ 146.56%No, they didn't magically create extra VRAM. 😁 That >100% number reflects effective reusable context under prefix-cache semantics, not literal physical KV capacity. The tool has already been tested against:🦙 llama.cpp ⚡ vLLM 🚀 NInfer 🔥 SGLang This is the benchmark I want to see more often. How much context is actually still there when the system is under load? Link to post in ALT.

On August 28th, at a military port in Qingdao, China, a detachment of the Northern War Zone Navy held the retirement ceremony of Ship 112 (ship name: Harbin Ship). All the officers and soldiers lined up neatly. Following an order, the “8.1-Army-flag” on the deck mast slowly lowered. The officers and soldiers presented a final military salute to Ship 112, which meant that Ship 112 had gloriously completed its mission mission and officially retired from the People’s Navy’s combat sequence. Ship 112 is the second-generation main warship that our country self-developed and built, and is the lead ship of the guided missile destroyer type 052. 8月28日，中国青岛某军港码头，北部战区海军某支队举行112舰（舰名：哈尔滨舰）退役仪式，全舰官兵整齐列队，随着一声令下，甲板桅杆上的八一军旗缓缓降下，官兵们向112舰献上最后一个军礼，这意味着112舰光荣完成了它的使命任务，正式退出人民海军战斗序列。112舰是我国自行研制建造的第二代主战舰艇，是052型导弹驱逐舰首舰。 #China# Navy

How well do L1s/L2s monetize per tx? And which one has the likeliest path to multi-billion dollar ARR? This is a good chart for understanding the trade-offs and use-cases of different L1/L2s as revenue-generating networks. I used Q2 2026 numbers, and for Base and Robinhood Chain I used the last 2 weeks, so Robinhood's numbers are most likely inflated and will probably sit closer to Base and Hyperliquid over time. Ethereum L1 is actually a wonderful business, you monetize at 30+ cents per tx, people are willing to pay for high-value/high-security tx but it doesn't scale beyond tens/hundreds of TPS. Solana sits at the other end of the spectrum: it monetizes at 0.52 cents per tx, but Solana can likely scale to 10k+ TPS, so it has more than 2 orders of magnitude more scaling headroom than ETH. Hyperliquid does similar TPS to L2s, close to hundreds of tx per second, but it obviously monetizes at the application layer and takes bips of transaction volume, so overall it brings substantial economics per tx. Hyperliquid says it can scale to 200k order/cancel messages per second, but from a realized-trades perspective it's probably in the low thousands, likely 10-25k TPS. L2s are actually pretty good businesses. Looking at Robinhood Chain over the last week, they monetize roughly 35+ cents/tx. Base is closer to single-digit cents per tx. Jesse has often talked about 1 gigagas/s as the ultimate goal, so if L2s can scale to that, it's roughly low thousands of TPS for complex DeFi transactions and 20-30k TPS for simple ETH transfers. So how many TPS do you need to become a multi-billion dollar ARR business? And command tens or hundreds of billions in market cap value? L1/L2 Monetization: TPS Needed for $1B / $5B / $10B / $25B ARR (at current revenue-per-tx run-rate) Robinhood Chain — 38.50¢/tx → 82 / 412 / 824 / 2,059 TPS Ethereum (L1) — 31.03¢/tx → 102 / 511 / 1,022 / 2,555 TPS Hyperliquid — 19.97¢/tx → 159 / 794 / 1,588 / 3,970 TPS Base — 1.44¢/tx → 2,202 / 11,010 / 22,021 / 55,052 TPS Solana — 0.526¢/tx → 6,029 / 30,142 / 60,285 / 150,712 TPS Looking at this table, Solana has a clear path at these low monetization rates to scale to multi-billion dollar ARR, it needs 60k TPS for $10B ARR. Ethereum currently doesn't scale to even high hundreds of TPS, so it's hard to see the L1 monetizing much further. L2s are great businesses for more centralized companies, it's a business that can clearly scale to $1B+ ARR, which is just huge for Robinhood or Coinbase long term in terms of earnings potential. Will be interesting to see how monetization rates per tx evolve over time. The big question for me: can we see an L1 scale to hundreds of thousands or 1M+ TPS at rates around single-digit cents per tx? Something reasonable.

Weekly Major Events on Calendar! 📅 May 18 🔵 $BEE unlocks 15.4M tokens (worth $0.38M, 27.992% of the circulation). 📅 May 19 🔵 $ZKJ unlocks 30.14M tokens (worth $0.36M, 5.976% of the circulation). 📅 May 20 🔵 $PYTH unlocks 2,124.67M tokens (worth $96.67M, 36.944% of the circulation). 🔵 $ZRO unlocks 25.71M tokens (worth $33.16M, 10.18% of the circulation). 🔵 $KAITO unlocks 12.11M tokens (worth $5.36M, 5.019% of the circulation). 📅 May 21 🔵 $GWEI unlocks 237.92M tokens (worth $36.09M, 13.595% of the circulation). 📅 May 22 🔵 $MBG unlocks 25M tokens (worth $7.88M, 10.087% of the circulation). 🔵 $HYPER unlocks 11.21M tokens (worth $1.18M, 3.314% of the circulation). 🔴 Final reading of U.S. one-year inflation expectations in May. 📅 May 23 🔵 $AVAIL unlocks 209.13M tokens (worth $0.84M, 5.468% of the circulation). 🔵 $ZORA unlocks 166.67M tokens (worth $1.9M, 3.728% of the circulation). 📅 May 24 🔵 $SOSO unlocks 15.82M tokens (worth $5.86M, 5.052% of the circulation).