Echoing sentiments shared in his “superintelligence”-focused blog post, Meta CEO Mark Zuckerberg expanded on his bullish ideas that glasses will be the primary way users interact with AI in the years ahead. During Meta’s second-quarter earnings call, the social networking exec told investors he believes people without AI glasses will be at a disadvantage in the future. “I continue to think that glasses are basically going to be the ideal form factor for AI, because you can let an AI see what you see throughout the day, hear what you hear, [and] talk to you,” Zuckerberg said. Adding a display to those glasses will then unlock more value, he said, whether that’s a wider, holographic field of view, as with Meta’s next-gen Orion AR glasses, or a smaller display that might ship in everyday AI eyewear. “I think in the future, if you don’t have glasses that have AI — or some way to interact with AI — I think you’re … probably [going to] be at a pretty significant cognitive disadvantage compared to other people,” Zuckerberg added. “The other thing that’s awesome about glasses is they are going to be the ideal way to blend the physical and digital worlds together,” he said. “So the whole Metaverse vision, I think, is going to … end up being extremely important, too, and AI is going to accelerate that.”
MIT’s algorithm combines ideas from algebra and the geometry into an optimization problem to build ML models with symmetric data using fewer data samples for training than classical approaches, resulting in improved accuracy and adaptability
A new study by MIT researchers shows the first method for machine learning with symmetry that is provably efficient in terms of both the amount of computation and data needed. These results clarify a foundational question, and they could aid researchers in the development of more powerful machine-learning models that are designed to handle symmetry. Such models would be useful in a variety of applications, from discovering new materials to identifying astronomical anomalies to unraveling complex climate patterns. “These symmetries are important because they are some sort of information that nature is telling us about the data, and we should take it into account in our machine-learning models. We’ve now shown that it is possible to do machine-learning with symmetric data in an efficient way,” says Behrooz Tahmasebi, an MIT graduate student and co-lead author of this study. They explored the statistical-computational tradeoff in machine learning with symmetric data. This tradeoff means methods that require fewer data can be more computationally expensive, so researchers need to find the right balance. Building on this theoretical evaluation, the researchers designed an efficient algorithm for machine learning with symmetric data. To do this, they borrowed ideas from algebra to shrink and simplify the problem. Then, they reformulated the problem using ideas from geometry that effectively capture symmetry. Finally, they combined the algebra and the geometry into an optimization problem that can be solved efficiently, resulting in their new algorithm. “Most of the theory and applications were focusing on either algebra or geometry. Here we just combined them,” Tahmasebi says.
By structuring multi-tranche notes with Fitch Ratings, LendingClub creates a risk-graded investment framework that appeals to institutional players like BlackRock, a game changer for fintech lending growth
The fintech lending landscape is undergoing a seismic shift as LendingClub Corporation and BlackRockcement a $1 billion partnership through 2026. This alliance, rooted in LendingClub’s LENDR (LendingClub Rated Notes) program, represents more than a capital infusion—it signals a strategic redefinition of how institutional investors allocate assets in alternative lending markets. For investors, this partnership underscores the growing institutional confidence in fintech-driven credit innovation and the potential for scalable, high-yield opportunities in a post-P2P era. LendingClub’s pivot from retail peer-to-peer (P2P) lending to institutional-grade platforms has been a masterstroke. The partnership’s $1 billion cap through 2026 reflects a long-term bet on LendingClub’s underwriting rigor and data-driven credit models. With over 150 billion data cells from repayment events across economic cycles, LendingClub’s predictive analytics now rival traditional banks’ risk assessment tools. This data edge, combined with structured finance expertise, positions the company to redefine institutional lending standards. The partnership’s significance extends beyond LendingClub’s balance sheet. It highlights a broader trend: institutional investors are increasingly allocating capital to alternative assets with higher risk-adjusted returns. BlackRock’s involvement in LENDR aligns with its Impact Opportunities (BIO) platform, which targets undercapitalized markets. By channeling funds into LendingClub’s loan portfolios, BlackRock gains exposure to a sector that balances financial returns with socio-economic impact—particularly in underserved consumer lending segments. This alignment is critical for investors seeking diversification. For fintechs, the LendingClub-BlackRock model sets a blueprint for scaling institutional partnerships. By offering multi-tranche structures with third-party credit ratings, fintechs can attract lower-cost capital from insurance companies, pension funds, and asset managers. This reduces reliance on volatile retail investors and creates a more stable funding pipeline. LendingClub’s $6 billion in loan sales since 2023—largely through structured certificates—demonstrates the scalability of this approach. Investors should also consider the strategic implications for LendingClub’s stock. The company’s Q2 2025 results—32% year-over-year loan growth, 33% revenue growth, and a 12% Return on Tangible Common Equity (ROTCE)—highlight its operational strength. With BlackRock’s $1 billion commitment, LendingClub can accelerate its digital banking initiatives, such as LevelUp Checking and Savings, further diversifying revenue streams. The partnership presents a dual opportunity: for LendingClub shareholders, the influx of institutional capital reduces funding costs, enhances profitability, and supports long-term growth; for institutional investors, LENDR notes offer a high-yield, credit-graded alternative to traditional fixed income, with downside protection via Fitch ratings. However, risks persist. Economic downturns could strain credit performance, and regulatory scrutiny of fintech lending remains a wildcard. Investors should monitor LendingClub’s net charge-off trends and BlackRock’s subsequent investment pace.
Nvidia’s new open, customizable reasoning VLM allows robots and vision agents to think about what they see similar to humans and plan about what’s in a scene using intelligence such as physics knowledge and common sense from training data
Nvidia Corp. is expanding its offerings of smarter AI models, physical intelligence for robotics and powerful enterprise AI servers. Nvidia unveiled that the Nvidia RTX Pro 6000 Blackwell Server Edition GPU, a graphics processing unit designed for servers, is now coming to enterprise servers. This new addition will allow organizations to run large language models at high speed and these 2U form-factor rack-mountable servers will use the Blackwell architecture to deliver high-performance AI inference workloads. The new Blackwell RTX Pro Servers bring GPU acceleration to traditional CPU-based workloads — including data analytics, simulation, video processing and graphics rendering — enabling up to 45 times better performance. According to Nvidia, this results in 18 times higher energy efficiency and significantly lower cost compared with CPU-only systems. Nvidia announced an expansion of its Nemotron model family, introducing two new models with advanced reasoning capabilities for building smarter AI agents: Nemotron Nano 2 and Llama Nemotron Super 1.5. These models deliver high accuracy for their size categories in areas such as scientific reasoning, coding, tool use, instruction following and chat. Designed to empower agents with deeper cognitive abilities, the models help AI systems explore options, weigh decisions and deliver results within defined constraints. Nemotron Nano 2 achieves up to six times higher token generation throughput compared to other models in its class. Llama Nemotron Super 1.5 offers top-tier performance and leads in reasoning accuracy, making it suitable for handling complex enterprise tasks. Nvidia announced Cosmos Reason, a new open, customizable 7 billion-parameter reasoning VLM for physical AI vision agents and robotics. It allows robots and vision agents to think about what they see similar to humans and plan about what’s in a scene using intelligence such as physics knowledge and common sense from training data. The company said it can help automate the curation and annotation of large, diverse training datasets, accelerating the development of high-accuracy AI models. It added that it can also serve as a sophisticated reasoning engine for robot planning, parsing complex instructions into steps for VLA models, even in new environments.
JPMorgan’s OpenAI coverage is just the start; with the US stock market at a record high, investment banks are keen to arrange secondary share sales for the hottest unicorns
Wall Street banks are starting to cover firms that are not publicly traded. JPMorgan Chase & Co. kicked off the trend with a report on OpenAI Inc. Citigroup Inc. followed suit a week later with a list of roughly 100 large private companies it will focus on, predominantly in the tech sector. It’s only a matter of time that their peers join the ranks. With the US stock market at a record high, investment banks are keen to arrange secondary share sales for the hottest unicorns. Brokerage fees aside, they will have direct access to startup employees who want to divest their stakes. Banks’ wealth managers, in turn, can offer services to these newly minted billionaires when they cash out. It’s a lucrative business for all. PMorgan’s research doesn’t offer price targets, but it does lay out a thought process. It’s instrumental to brokers who have largely relied on media reports to educate their clients. OpenAI has reportedly told investors that it could hit $174 billion in revenue by 2030, up from an estimated $13 billion this year. Since Big Tech are on average valued at 10 times sales, OpenAI could be worth $1.7 trillion in five years’ time, thereby giving substantial upside to those who buy into secondary sales. JPMorgan’s analysts are throwing cold water on this logic, calling the 2030 sales projection “ambitious.” To hit this target, OpenAI would need to capture about a quarter of the total market share, which would require “flawless execution” since it’s still at the early stage of commercialization. Rather, the focus should be on how fast the startup can achieve scale. If, say, OpenAI could generate $100 billion in revenue faster than Meta Platforms Inc., which achieved the goal in 17 years, that would be quite an unprecedented feat. Meta is a $2 trillion company. For once, this kind of thematic research provides direct value-add to institutional investors who increasingly find brokerage reports irrelevant. Market concentration means that asset managers just have to understand a few dozen stocks, thereby reducing their need to outsource due diligence to investment banks. This new trend therefore provides a chance for analysts to showcase their worth. It will be more work, however. Instead of writing about quarterly earnings, which more or less read the same over time, they will have to talk to customers, business partners and everyone else within the ecosystem to understand billion-dollar tech companies that may never go public.
KeyBank’s move to UJET’s Google Cloud–based CCaaS platform saw agent call volumes decrease by 15% while seeing a 50% increase in agent digital chat volumes and reduction in costs to run the contact center by 10%
As KeyBank continues migrating its systems to the cloud, it has begun to see results. The Cleveland-based bank fully transitioned its contact center to UJET’s Google Cloud–based CCaaS platform, retiring multiple legacy systems. The move led to a 15% decrease in agent call volumes, a 50% increase in digital chat volumes, and a 10% reduction in costs. “We’re rolling out new capabilities that our contact center teammates are loving … it’s going to simplify the role for our teammates, and when we do that we ultimately are going to deliver a better experience for our clients,” said Jordan Olack, director of intelligent automation and contact center. Cloud adoption is accelerating across the industry: Capital One has operated fully on AWS since 2021, JPMorgan Chase partnered with Thought Machine, Wells Fargo opted for a dual-cloud strategy, and Citi announced in 2024 it would migrate some apps to Google Cloud. KeyBank began its own transition in 2019, later expanding to infrastructure in 2022, with the UJET migration spanning May 2023 to October 2024. “We’re actually going right to the source … We are a strategic partner of UJET, and we’ve made equity investments as a bank into UJET as well,” Olack said. “We’ve never thought of them as a ‘legacy bank,’ but rather as an established leader who was looking for a technology partner that could match their ambition,” said UJET CEO Vasili Triant. He added: “Bridging the gap between on-premise procedures and a cloud-native mindset required a highly collaborative and conversational process.” Dylan Lerner, senior analyst at Javelin Strategy, emphasized that cloud transformation is gradual: “Banks rarely go all-in on cloud technology … A piecemeal approach is easier to manage and gives the financial institution time to adjust.” At Cenlar, AI is being applied to reshape mortgage servicing. EVP and COO Leslie Peeler said her IBM experience showed how AI could compress weeks of work into minutes: “That really informed my thinking around how AI could be disruptive beyond smaller point solutions by stringing things together.” “There’s a lot of cost opportunity, certainly, to use AI and generative AI to take cost out of servicing,” Peeler noted, citing chatbot pilots that sped up homeowner responses. Cenlar, which services over 2 million loans, partnered with PhoenixTeam on AI compliance tools: “The process … takes that from a five- to six-week effort down to one to two days.” She stressed that servicing can now rival origination in technology focus: “I don’t think servicing needs to take a back seat.” Cenlar is also launching a “value network” giving managers cockpit-style visibility. “It’s the equivalent of moving from a lack of visibility to what’s under the water to complete visibility,” Peeler said.
Citi’s new retail head of retail banking says simplified strategy distinguishes Citi from larger banks; credits fintechs with pushing traditional banks to reduce complexity
Kate Luft, Citi’s new head of U.S. retail banking, draws inspiration for customer engagement from the airline industry. “I think of it like an airline,” she said. “The more you do with us, the more we recognize you.” Luft led the overhaul of Citi’s U.S. retail strategy last year, simplifying products, consolidating checking accounts, and introducing “relationship tiers” that reward customers with perks like waived fees when they meet balance thresholds. “Really what we did was redefine our products and value [propositions],” she said. “Our mandate was, how do we make it super-simple for our clients?”
Treasury and Trade Solutions (TTS) and embedded finance platforms driving growth for banks in cross-border transactions and deposit balances amid operational uncertainty, FinTech fragmentation and growing demand for streamlined, data-rich payments
Financial tools once limited to large firms are now accessible to SMBs via APIs and embedded finance. From large lenders like Citi and JPMorgan, to systemically important banks like BNY and Lloyds, as well as major market institutions like Truist, bank executives all stressed to their respective investors the importance of back-office units. Against a backdrop of operational uncertainty, FinTech fragmentation and growing demand for streamlined, data-rich payments, these FIs’ TTS and embedded finance platforms are becoming strategic growth engines. Done right, embedded finance shifts from a buzzword to a reliable infrastructure. Citi’s Services business, for example, posted record second-quarter 2025 revenues of $5.1 billion, up 8% year over year. Market share gains of 40 basis points in TTS were driven by a 7% rise in cross-border transaction value and higher deposit balances. BNY’s Treasury Services offerings were likewise up year-over-year. Nearly two-thirds would switch providers to access embedded finance solutions. For banks, this underscores the opportunity within treasury and payments services. At the same time, innovations like stablecoins are reshaping what treasury management and payments might look like in the future.
A stablecoin is a blockchain token backed 1:1 by cash or cash-like assets, used as a substitute for fiat in on-chain trade with use cases in trade settlement, remittances and online purchases, whereas tokenized deposits are bank-issued tokens backed by dollars held in client accounts
With the GENIUS Act now law, U.S. banks are expected to increasingly explore issuing blockchain-based assets. While many tout stablecoins for faster, cheaper payments, most banks are actually eyeing tokenized deposits — not stablecoins — as the more viable product. Though both are digital tokens tied to fiat value, their nature and implications differ greatly. Stablecoins (like USDC) are backed 1:1 by cash or equivalents, circulate on public blockchains, and are used broadly as money for trade, savings, and remittances. Tokenized deposits, by contrast, are representations of client bank deposits, issued and moved within a bank’s private network, with value transfers still tied to bank-controlled ledgers. Unlike stablecoins, they’re non-fungible across institutions and don’t circulate freely. Their purpose is to modernize existing bank services, not create new monetary systems. The difference lies in function and intent: stablecoins are a new form of money, while deposit tokens are tools for enhancing traditional banking. As banks increasingly mention blockchain innovations, it’s vital to distinguish between the two.
SEC’s Atkins says most crypto assets are not securities; plans purpose-fit disclosures for crypto securities including for so-called ‘initial coin offerings,’ ‘airdrops’ and network rewards.”; could allow innovation with ‘super-apps’
SEC Chairman Paul Atkins said his agency is launching “Project Crypto” with an aim to make a quick start on the new crypto policies urged by President Donald Trump. Atkins said the effort will be rooted in the recommendations of the President’s Working Group report issued Wednesday by the White House. He described it as “a commission-wide initiative to modernize the securities rules and regulations to enable America’s financial markets to move on-chain.” “I have directed the commission staff to draft clear and simple rules of the road for crypto asset distributions, custody, and trading for public notice and comment,” Atkins said. “While the commission staff works to finalize these regulations, the commission and its staff will in the coming months consider using interpretative, exemptive and other authorities to make sure that archaic rules and regulations do not smother innovation and entrepreneurship in America. Despite what the SEC has said in the past, most crypto assets are not securities,” Atkins said. Atkins suggested his agency will move to begin answering those questions now, working on “clear guidelines that market participants can use to determine whether a crypto asset is a security or subject to an investment contract.” For crypto securities, he said he’s “asked staff to propose purpose-fit disclosures, exemptions, and safe harbors, including for so-called ‘initial coin offerings,’ ‘airdrops’ and network rewards.” Atkins said he means to “allow market participants to innovate with ‘super-apps'” that offer a “broad range of products and services under one roof with a single license.”