I just clocked the iPhone 18 Pro pulling 45W from my power bank. That's progress. I promise you there are people who will still plug in a 5W adapter or an 18W adapter from when Apple gave them to you for free.

Gadget of the day. A Bluetooth speaker shaped like a retro tape deck. 2+2 acoustic design, 5W+5W output, LDAC high-res Bluetooth, USB DAC mode, 10.5 hours of battery life, and HiFi/Retro sound presets. Compatible with the JM21, M21, and M33 players. The FIIO Retro Box...

Last wk, SPX/Nas/R2K +0.5%/+0.8%/-1.5% w/ oil -4%. But a hawkish Warsh on Friday led to a bear flattening of the yield curve. Despite $NVDA guide of 70% CY27 rev growth vs consensus of 47%, SOX Index -2.3% while software $IGV +5.9% on solid earnings. In general, many AI investors have been bullish on semiconductors and bearish on software on the belief that AI will displace many point solution software companies. This is why the SOX index is up 62% YTD and IGV is still only up 4% YTD versus the S&P +13%. Situational Awareness was the poster child for this type of positioning. But since the unwinding of the Momentum trade which started on 6/22 (I wrote about these concerns on 6/20), IGV has rallied 25% while the SOX Index has declined 22% through 8/28. For perspective, the Morgan Stanley Momentum index (momentum long performance minus momentum short performance) from 6/22-8/28 is down 36% while their more concentrated TMT index is down 54%. But a bullish twist on AI for the software sector introduced recently is that AI agents will access software tools ~10-100x more often than humans. On 8/6, $TEAM, which was in the bucket of software names widely considered at risk of being replaced by AI, rallied 35% the next day in reaction to solid earnings & outlook. Then on 8/13, $WDAY rallied 18% on the news that private equity firm Silverlake might be pursuing an acquisition which I wrote probably put a floor underneath software. Workday was also supposed to be in the AI crosshairs and private equity has higher bars to clear given their use of leverage and holding period than a typical investor. Then on 8/26, $CRM reported solid results, guidance and a deal with Anthropic (in which they also first invested in May of 2023.) The stock was up 23% in reaction the next day. This seemed to be a strong counterpoint to the SaaS-pocalypse worries. This strategic alliance allows users to execute actions natively inside Claude without needing to open traditional software screens. Salesforce also seems to be changing how they charge customers with fees more related to customer use and benefits to their business. Then on 8/27, Workday reported results which were good enough but arguably acquisition prospects drove more of the stock reaction of +6% the next day from the slightly down opening price. Historically, system of record, security and gaming software have been the only three areas I have liked within software. I now wonder whether the fundamental implications of Atlassian, Workday and Salesforce are supportive of the technical reactions in the software stocks as a group as agentic AI continues to ramp. So how do I square this with my concerns that the rapidly escalating amounts spent on AI by corporations has to come from somewhere? Annualized revenue run-rates for Anthropic and OpenAI have ramped from $29B to start the year to $105B just 7 months later. Software spending globally excluding AI was roughly $1 trillion in 2025. But IT services at $1.7 trillion is a bigger category which I believe still has risk. And finally, knowledge worker compensation is an even bigger category where disruption would be even less noticeable at an estimated $35-50 trillion in 2025 or roughly 30% of the global workforce. Looking forward, the deal on Friday for Venezuelan oil fields that hold the largest crude reserves in the world at 17-18% should get us off to a positive start to the week with declining oil prices. But a bit further out: 1) “Don’t Fight the Fed” given I believe a hike is likely on 9/16 because the 10/28 mtg is right before mid-terms, 2) September has the poorest seasonality of all months, 3) there is even worse seasonality than normal during mid-term election years (see prior posts for more detail) and 4) recent bipartisan pushback against datacenter expansion (one of the few things both sides seem to agree on though I believe this is wrong and hope it will change with more education) puts pressure on the AI infrastructure names. As Warren Buffett says, the market has to keep pitching but you do not need to swing.

Haha! This is crazy. Maestro v1.8.5 w/ MiniMax-Music3 just dropped. As a test, I prompted "Wheels on the Bus as a death metal song by the band Nursery Crimes." But it's Maestro, so I told it to auto-make the entire video using MiniMax-H3 for video.🤯 Free, Local & Open! 🔊🎵📽️

Digital Credit - Weekly Close (Sep. 18) $BTC broke $81K and digital credit issuers’ common equities ran with it, big time… 🚀 @Strategy is back to compounding... • $MSTR +17.5% w/w to $153.92, its best week since August 2026. Net mNAV rebuilt to 1.19x and the treasury now sits on $4.9B of unrealized gains. 😱 • $STRC $98.51 (-0.1% w/w), 12.2% effective yield. Another $139M repurchased announced earlier in the week; the buyback program is approaching $1B retired while $STRC climbs back to $100 @Strive hit a massive milestone... • $ASST +8.9% w/w to $30.09. Net mNAV 2.37x, and $BTC yield YTD is up to +49.5%. 🤯 • $SATA back above par at $100.01 (+0.3% w/w), 13.0% effective yield. The treasury crossed 25,000 $BTC, and what is left of the short pays 28.1%/yr with half a day to cover. Where digital credit closed, updated daily. 👇

We entered AIRA₃ with an ensemble of models in the live competition, and also assessed it with several others post-hoc. The 8th ranked gold medal entry ensemble was a combination of GPT 5.5 (w/ OpenCode) + Claude 4.8 (w/ ClaudeCode). Post-hoc we assessed with Muse Spark 1.2 (w/ MuseCode), which also performed at a Gold Medal level, as well as Muse Spark 1.1 (w/ OpenCode) and GLM 5.2 (w/ OpenCode), both of which achieved Silver Medal level performance. The post-hoc submissions were also graded externally on the same private test set as those made during the live competition.

Will Embodied AI Be AI’s Next Breakout? At this week’s World Humanoid Robot Games, organizers released a dataset containing more than 2,500 hours of real-world robot operation data. It reflects a broader shift: embodied AI is moving beyond stage demos and toward data accumulated through physical interaction. Zhihu contributor 任广杰, Assistant Vice President at @LejuRobotics_ , believes embodied AI will become AI’s next breakout field. But it will not resemble ChatGPT’s overnight rise. The real signal will be a steady increase in the hours robots can work reliably in real environments. 1️⃣ Physical AI plays by different rules The previous AI waves, including computer vision and NLP, operated largely in information space. When a model makes a mistake, the task can usually be retried. Robots operate in the physical world. Changes in lighting, friction, payload, or object position can damage equipment or halt a production line. That creates two requirements AI has rarely had to meet at scale: 🔹 Real-time, closed-loop control 🔹 Consistently high success rates in uncontrolled environments A successful demo is not enough. The system must continue working when conditions change. 2️⃣ Three inflection points have arrived 🔹 Robot bodies are becoming reliable enough Integrated actuators now combine motors, reducers, encoders, and drivers. This has sharply reduced joint failure rates. At Leju, robot mean time between failures has increased from hours to thousands of hours. Below that threshold, the economics do not work. Maintenance and downtime can erase the value of replacing human labor. 🔹 Robots now have both a “brain” and a “cerebellum” Traditional industrial robots execute fixed instructions. A change in task usually requires new programming and teaching. Large models give robots task-level understanding. A robot can inspect randomly placed components, then plan the order and pose for grasping them. Meanwhile, lower-level control systems handle balance, locomotion, and physical disturbances. The “brain” decides what to do. The “cerebellum” keeps the movement stable. The body must survive repeated execution. 🔹 The data flywheel is starting to turn Embodied AI is beginning to follow the path of autonomous driving. Real robots generate operational data. That data improves the models, which then return to the physical world and generate more useful experience. This closed loop is essential because many problems only appear during deployment. 3️⃣ The real gap lies between the lab and the factory Leju began developing humanoid robots in 2016. One of the clearest changes in the current cycle is that factories are now willing to open real production stations for testing. Three years ago, that was much harder. A 95% grasping success rate may look impressive in a laboratory. A factory may require 99% availability under changing lighting, reflective surfaces, deformed containers, and shifting object positions. Customers therefore ask very practical questions: 🔹 How long can the robot operate without failure? 🔹 Can it recover from an exception by itself? 🔹 Can it keep pace with the production line? Answering them requires the body, controller, model, data loop, and deployment engineering to work as one system. 4️⃣ The first commercial wave is already taking shape The earliest deployments appear in structured environments with clear task boundaries and measurable value. Leju’s full-size Kuavo 5 robots are already used for reception and guided tours in exhibition halls, banks, and stores, as well as for park inspection. The company says these projects now cover 23 Chinese provinces and hundreds of customer cases. In manufacturing, Leju explored more than 70 factories and hundreds of potential tasks over the past year. Dozens passed proof-of-concept acceptance. Its wheeled Kuavo 5-W has entered small-batch deployment for tasks such as loading small automotive components and depalletizing cartons. The important change is that some customers are beginning to pay for repeatable workflows, not merely demonstration projects. 5️⃣ Home robotics remains the endgame Homes may eventually become the largest market, but they are also the hardest environment. They are highly unstructured. Task boundaries are almost unlimited, and consumers are extremely sensitive to cost. Two conditions must be met: ✅ Embodied models must generalize well enough for real household use. ✅ The total cost of the robot must fall below the cost of human labor. Until both happen, industrial and commercial settings will remain the more realistic path to scale. 6️⃣ Watch effective working hours, not flashy demos The best measure of an embodied AI system is how long it can create value without human intervention. Three questions matter: 🔹 Can it do the job? This depends on whether the task can be decomposed and whether the environment is sufficiently structured. 🔹 Can it do the job reliably? The key metrics are continuous operating time and autonomous recovery from failures. 🔹 Is it economically worthwhile? The customer must be able to calculate the full lifecycle cost. Once the economics become clear, repeat purchases can begin. Embodied AI is already moving from exhibition piece to production tool. Its breakout will not be defined by one viral robot. It will be defined by thousands of machines quietly accumulating useful working hours in factories, stores, and eventually homes. 🔗 Full analysis: #EmbodiedAI# #HumanoidRobots# #Robotics# #PhysicalAI# #ArtificialIntelligence# #ChinaTech#