Register and share your invite link to earn from video plays and referrals.

Search results for NU_Point_Of_View
NU_Point_Of_View community
One keyword maps to one global community path.
Create community
People
Not Found
Tweets including NU_Point_Of_View
A TON OF THINGS HAPPENED IN THE STOCK MARKET TODAY. Here's a full recap: 1. SpaceX $SPCX rose about 7.6% on Monday to roughly $171.09 after Elon Musk confirmed that discussions with TSMC are real, though still in the early stages. The talks follow reports that TSMC is exploring a role in Terafab, the Texas chip venture tied to Tesla, SpaceX, and xAI, which could eventually invest $119B to manufacture chips for products including Optimus, Cybercab, and future AI infrastructure. The prospect of TSMC participating was viewed as strengthening SpaceX’s long-term AI chip supply strategy and potentially reducing Intel’s future role in the project, helping send Intel shares lower while TSMC reached a new all-time high. A reiterated Morgan Stanley Overweight rating and $300 price target provided additional support for the stock. 2. Meta $META and Microsoft $MSFT are reportedly reducing internal use of Anthropic’s Claude as they increasingly shift employees toward their own AI tools, according to The Information. Microsoft had been on pace to spend at least $1B annually on internal Claude usage earlier this year but has since reduced that level by more than one-third. At Meta, the number of Claude Code users has reportedly fallen to about 30,000 from roughly 60,000 earlier this year, while MetaCode now has more than 30,000 users and Muse Code has surpassed 6,000. Despite the shift, Meta still spent more than $105M on Claude Code over a recent 28-day period. 3. JPMorgan expects the memory market $MU $SNDK $WDC $SKHY $STX to remain supply-constrained through at least 2028, with customer order discussions already extending as far out as 2031. The firm forecasts HBM bit demand to grow 63%, non-HBM server DRAM demand to increase 37%, and CY2027 HBM blended ASPs to rise 54% year over year. JPM also believes Micron’s expectation of having more than 35% of revenue covered by long-term supply agreements through 2030 may prove conservative, noting that several Asian memory manufacturers already have 50%+ of their capacity secured under similar long-term agreements. 4. RBC Capital Markets reiterated its Outperform rating on Zeta $ZETA and raised its price target to $40 from $31, implying roughly 22.6% upside. Analyst Matt Swanson said recent investor meetings highlighted growing momentum around strategic partnerships and adoption of Athena, while emphasizing that RBC does not believe these opportunities are materially reflected in either current guidance or the company’s 2030 outlook. RBC also argued that Zeta has increasingly differentiated itself from traditional ad-tech peers and is now being valued more like a software company. The latest increase marks RBC’s third price target raise on Zeta this calendar year. 5. Ondas $ONDS secured a $56M order for Electronic Safe & Arm Devices (ESADs) supporting a European loitering munition program. The contract will be fulfilled through Ondas' recently acquired Bron Technologies and GATE Technologies businesses. ESADs manage a munition’s safety, arming, and initiation functions, with production set to move forward using technology that has already been qualified for deployment. The award further expands Ondas’ presence in the growing European defense and precision munitions market. 6. Bank of America, Citi, and Morgan Stanley have begun syndicating a $60B debt package tied to Anthropic’s lease of Google-designed TPUs supplied through Broadcom $AVGO, making it the largest chip-financing deal to date. The package includes $42B of senior secured debt backed by Broadcom and $18B of junior debt without Broadcom guarantees, with Blackstone already committing roughly $9B to the junior tranche. The financing is expected to fund Anthropic’s 2027 chip orders, with Broadcom’s support helping reduce borrowing costs and potentially allowing the senior debt to be refinanced into the investment-grade market. The deal centers on Google $GOOGL TPUs, developed with Broadcom as an alternative to NVIDIA $NVDA GPUs for large-scale AI workloads, while Broadcom filings also indicate Anthropic could issue up to $42B of convertible notes to Broadcom to cover future lease payments. 7. U.S. Strategic Petroleum Reserve (SPR) inventories fell another 767,000 barrels last week to 283M barrels, the lowest level since 1982. The decline marks the 28th consecutive weekly drawdown, leaving the SPR down nearly 133M barrels from its recent peak. As a result, America's emergency crude oil reserves continue to sit near their lowest levels in more than four decades, underscoring the limited strategic buffer available despite elevated geopolitical risks. 8. The Consumer Confidence Index fell 6.7 points in September to 81.9, its lowest level since April 2014, marking a fourth consecutive monthly decline and an 11.2-point drop over that period. The Present Situation Index declined 7.9 points to 109.3, its weakest reading since February 2021, while the Expectations Index fell 5.9 points to 63.6, its lowest since April 2025 and a third straight monthly decline. Confidence has weakened across all age groups and nearly all income brackets over the past six months. Consumers increasingly cited high prices, the cost of goods and services, and oil and gasoline prices as the primary factors weighing on the economic outlook. 9. Brazilian markets rallied after Flávio Bolsonaro finished ahead of President Luiz Inácio Lula da Silva in the first round of Brazil’s presidential election, with investors reacting positively to the prospect of a more market-friendly policy agenda ahead of the October 25 runoff. Brazilian assets including Nu Holdings $NU, the iShares MSCI Brazil ETF $EWZ, and other Brazil-exposed equities such as MercadoLibre $MELI outperformed following the result. The market reaction reflected optimism around potential fiscal and economic policy changes, though the election remains undecided with a runoff still to come. 10. The top 10 most active options today by contracts traded were $NVDA with 4.0M contracts, $TSLA with 3.2M contracts, $SPCX with 1.7M contracts, $AAPL with 1.1M contracts, $MSFT with 874K contracts, $MU with 872K contracts, $AMZN with 844K contracts, $INTC with 835K contracts, $META with 790K contracts, and $GOOGL with 502K contracts. 11. OpenAI is reportedly in talks with UAE investment funds, including MGX, and BlackRock $BLK to anchor a new $30B+ funding round at a valuation of roughly $1.4T, according to Bloomberg. UAE investors have discussed contributing as much as $10B collectively, while existing backers Thrive Capital and Andreessen Horowitz are also considering additional investments, alongside discussions with the University of California endowment. Rather than negotiating with a lead investor, OpenAI is reportedly presenting investors with a fixed ~$1.4T pre-money valuation. The fundraising comes as OpenAI continues to delay an IPO until at least 2027 while raising unprecedented amounts of capital to fund AI model development and large-scale infrastructure expansion. 12. Melius upgraded Microsoft $MSFT to Buy with a $665 price target, arguing that growing concerns around AI safety are actually strengthening Microsoft's competitive position. The firm said enterprises are increasingly choosing to spend on a secure AI wrapper that can route models to the right tasks, manage AI agents, and provide governance rather than relying directly on frontier AI labs. Melius believes Satya Nadella and Microsoft's leadership are becoming the "adults in charge" for AI security and governance, and now expects Azure revenue growth to accelerate to above 50% by F4Q27. WALL STREET IS THE GREATEST SHOW ON EARTH.
Show more
0
71
2.2K
110
Forward to community
Woah! @Nubank just got authorization to become a bank in Mexico. 15% of Mexican adults were already Nu customers before it was legally a bank. The numbers from the press release: - 15 million customers, adding 12,000 per day (!!!) - $5.9bn in deposits - 54% of customers got their first-ever credit card from Nu - Breakeven in Q1 2026, with the efficiency ratio improving 78 percentage points - A projected $4.2bn investment in Mexico through 2030 Nubank is a monster. Just a monster. And the detail that got me: Nu built all of this on a SOFIPO license, a wrapper originally designed for community savings co-ops. The company calls the conversion "unprecedented" because no SOFIPO has ever transformed into a bank before. Nu is the first. Which makes Mexico arguably the most competitive digital banking market anywhere right now. Revolut launched its first bank outside Europe there in January. Plata (built by the ex-Tinkoff team, now a $3.1bn valuation) got its de novo license in February. Ualá and Klar bought theirs. Santander launched Openbank. Hey Banco spun out of Banregio. Every region's neobank war produced a winner, and those winners have never fought each other on the same turf. Brazil's champion, Europe's most aggressive fintech, the Tinkoff diaspora, Argentina's Ualá. Mexico is the first market where all the playbooks collide. Nubank enters that fight with 15 million customers, a deposit book, and breakeven economics. What I wonder about is the next product. A bank license unlocks nómina, and payroll is how Mexican banks lock customers in. Whoever wins payroll wins the market. And Nubank's take on that could be fascinating.
Show more
What if Your Neural Network Was Forced to Obey Physics? Physics-Informed Neural Networks (PINNs) are neural networks trained to satisfy a differential equation by building the PDE residual directly into the loss. They emerged from a very practical problem...classical PDE pipelines can be brilliant, but they often demand heavy discretization work (meshes, stencils, stability tuning), and the method you build is usually tied to one geometry and one solver setup. A PINN flips the workflow by representing the solution itself as a smooth function uᵩ(x,t) and enforcing the physics everywhere you choose to sample the domain. People often meet PINNs in the least helpful way...via a flashy solution plot, and almost no explanation of what was enforced to get it. In this series we keep the enforcement visible. We pick a differential equation, represent the unknown solution as a flexible function, measure how well that function satisfies the equation across the domain, and train it to reduce that mismatch everywhere we sample. A normal neural net learns from labels...you give it inputs and target outputs. A PINN learns from a differential equation...you give it inputs (x,t) and it gets punished whenever its output fails the PDE. By punish we mean that the loss increases when the mismatch is large we reward it if the loss decreases as the mismatch gets smaller. The network isn’t replacing physics, it’s becoming a flexible function that is forced to satisfy the same calculus you’d impose on any candidate solution. The math breakdown: We start with a PDE we want to solve on a domain Ω. Write it as uₜ(x,t) + N(u(x,t), uₓ(x,t), uₓₓ(x,t), …) = 0 for (x,t) in Ω A PINN replaces the unknown function u with a neural network output uᵩ(x,t) Now define the physics residual by plugging uᵩ into the PDE rᵩ(x,t) = ∂uᵩ/∂t + N(uᵩ, ∂uᵩ/∂x, ∂²uᵩ/∂x², …) If uᵩ were an exact solution, we would have rᵩ(x,t) = 0 everywhere. We may also have data points (xᵢ,tᵢ,uᵢ) from measurements or a known initial condition. The training objective is just a weighted sum of squared errors L(ᵩ) = L_data(ᵩ) + λ L_phys(ᵩ) + L_bc/ic(ᵩ) with L_data(ᵩ) = meanᵢ |uᵩ(xᵢ,tᵢ) − uᵢ|² L_phys(ᵩ) = meanⱼ |rᵩ(xⱼ,tⱼ)|² where (xⱼ,tⱼ) are the collocation points in Ω L_bc/ic(ᵩ) = penalties enforcing boundary conditions and initial conditions The key technical step is that the derivatives inside rᵩ are computed by automatic differentiation ∂uᵩ/∂t, ∂uᵩ/∂x, ∂²uᵩ/∂x², … So we can differentiate the total loss L(ᵩ) with respect to ᵩ and train with gradient descent. This is the whole idea behind PINNs. Learn a function, but make the PDE part of the loss, so the network is trained to be a solution, not just a curve-fitter. In the render, the main 3D surface is the network’s current guess uᵩ(x,t), drawn as a living sheet over the (x,t) plane. Hovering above is the neural scaffold...a visible graph of feature nodes and connections. The bright tension threads are the physics residual rᵩ(x,t): each thread tethers a collocation bead on the sheet up to the scaffold, and it thickens and brightens exactly where |rᵩ| is large (color encodes the sign). As training runs, those threads go slack across the domain not because we hid the error, but because the network has actually been pushed toward rᵩ(x,t) ≈ 0. #PINNs# #PhysicsInformedNeuralNetworks# #ScientificMachineLearning# #PDE# #DifferentialEquations# #Optimization# #MachineLearning# #AppliedMath# #ComputationalPhysics#
Show more
How does an AI coach that can't touch medicine prove it's safe? Numan, a UK digital health company, built Nu, an AI coach that is a wellness product: no diagnosis, no prescriptions. It had to hold that line while still coaching well. The team first scored replies one by one on static datasets, but averages hid a few dangerous replies, and scripted conversations couldn't follow Nu's unexpected turns. Worse, some low scores came from a rubric demanding empathy and advice within two turns. The turning point was treating evaluation criteria, not prompts, as the scarce asset. They test the exact same LangGraph graph patients use, score by theme, let safety probes block regressions at release, and have a cheaper model play patients so a judge can score whole conversations. A provider migration that usually takes weeks took about a day. How Numan built an eval-first AI health coach with LangGraph and LangSmith #LangGraph# #LLMEvaluation#
Show more
Nu Holdings named Estephany Paulette Ley as chief executive officer of its Mexico operation, part of a broader leadership reshuffling of the Brazil-based digital bank
Nu checked the entry paths across five of Dr4v3n's contacts. No repeated node sequences, no repeated timing intervals. The variation was too consistent for randomness and too perfect for improvisation. The pattern was deliberate. @FinalbosuX
Show more
NU Silver Arrows Radio Show, Episode 3: Montreal Unpacked Go behind the scenes of the Canadian Grand Prix as the Mercedes-AMG PETRONAS F1 Team breaks down a weekend of highs, lows, and everything in between. Watch Episode 3 now on the team’s YouTube channel
Show more
Nu Holdings - Q1 2026 Earnings $NU 11.7 [-9.5% AH] ⚪ Revenue: $5.0B (Est: $5.0B) ✅ Adj. EPS: $0.21 (Est: $0.19) Additional Metrics: Net Income: $871M [+41% YoY] ROE: 29% Net Interest Income: $3.25B [+12% QoQ] Total Customers: 135M+
Show more
What nu metal band's name is a misspelled vegetable? 🌽 Play the Billboard Daily Crossword now ➡️
AT Nu Udra is really fun!! Once you know what to do its very great hunt Btw for lance, the charge counter timing vs the nuke attack is when he dashes backwards with his fire, release the counter immidietly when he slides away and you can nullify the entire thing ☺️
Show more