독서

Morgan Stanley published a video interview on October 19, 2026, featuring Carolyn Campbell, the firm's Asset-Backed Securities (ABS) Strategist. The episode focuses on the specific impact of the current macro-financial environment on the expansion of artificial intelligence infrastructure, particularly the role played by changes in borrowing costs.

According to the official video description, Campbell explains how rising borrowing costs are currently slowing down the AI data center boom. She also analyzes which conditions or market dynamics could serve as key drivers for a potential rebound in financing activities for such projects.

For professional readers interested in the capital structure of the AI industry, debt financing channels, and the return cycles of infrastructure investments, this video offers an immediate perspective from a frontline strategist at a major investment bank. While a full transcript is not provided, the clear topic and speaker identity make it a valid resource entry point for understanding current bottlenecks and opportunities in AI financing.

At the AI Engineer World's Fair 2026 in San Francisco, Pamela Fox, Principal Cloud Advocate at Microsoft, conducted a workshop on model swapping. She argued that generic benchmarks often fail to reflect how large language models actually behave in specific applications, urging developers to swap and test models directly within their own scenarios.

Fox compared GPT-5.5, Kimi K2.6, DeepSeek V4 Flash, and Mistral Large 3 on the Microsoft Foundry platform. Her demonstration progressed from basic letter-counting prompts to complex tasks including Retrieval-Augmented Generation (RAG) groundedness, image input handling, multilingual pitfalls, tool calling, and building a travel agent with LangChain v1.

Beyond functional testing, she detailed evaluation methodologies such as string-check evals, LLM-as-a-judge, using Microsoft's new ASSERT tool to generate test sets from specifications, and optimizing prompts with DSPy to find instructions that even the cheapest model can satisfy. This sequence provided engineers with a practical workflow for model selection and evaluation.

The Macro Mondays podcast, part of the Real Vision network, released a new episode titled "Is Inflation About to Derail the Rally?" on October 19, 2026. This week's episode features Andreas Steno, founder of Steno Research and former Global Chief Strategist at Nordea Bank, flying solo to provide actionable views on macro and market outlooks.

According to the show notes, Andreas Steno breaks down the latest inflation signals and their implications for the market. With energy prices and input costs rising, he discusses whether another inflation shock is on the horizon and poses a critical question: could Kevin Warsh be focusing on the wrong inflation measure? He also examines the latest business cycle data and explains why he is not yet ready to turn bearish on equities.

Regarding AI industry investment topics, the episode timestamps indicate discussions on several relevant subjects, including "AI Credit Stress: Is the Data Center Buildout Finally Breaking?" and "Are the Mag 7 Really About to Cut AI CapEx?" Steno also touches upon how rising compute prices change the AI bear case and the potential for a new AI trade war following DeepSeek's ban from US government systems.

Beyond AI and inflation, the episode covers European market dynamics, specifically the renewed pressure in France's bond market and its potential meaning for European investors. For industry readers focused on macro liquidity, AI infrastructure financing risks, and geopolitical impacts on tech supply chains, this podcast offers a comprehensive discussion framework from a seasoned strategist's perspective.

Meta's ad recommendation team published an engineering blog post describing a method that uses semantic tokens to replace traditional ad IDs, reporting that the approach, paired with Meta's Generative Ads Model (GEM) and Adaptive Ranking Model, delivered a 0.35% lift in conversions across surfaces. The data comes from Meta's internal measurements in February 2026.

Traditional recommendation systems assign each ad, product, or page a dedicated ID, and the model can only learn from that ID's historical engagement. Semantic tokenization works differently: a pre-trained multimodal model reads an ad's text and images, encodes it into an embedding carrying world knowledge, decomposes that embedding into independent subspaces, and quantizes each subspace to a discrete token from a codebook. Similar items share tokens, much as related words share roots.

This means a newly launched running shoe ad does not need to accumulate engagement data from scratch—it inherits the category knowledge the model already learned from previous-season running shoe ads. Meta's team uses trail shoes as an example: when a sportswear brand launches next season's trail shoe, tokenization lets the new product and its ads build on what the model learned from last season's shoe, even before anyone sees the new ad.

The most practically valuable finding in this engineering post is not the conversion number itself but the evaluation method. Meta tested eight versions of the tokenizer and found a strong correlation between intrinsic quality metrics—cluster purity (whether items sharing a code tend to belong to the same category), distribution balance (whether items spread evenly across semantic IDs rather than crowding into a few), and codebook utilization (whether a larger share of codes holds at least one item)—and downstream ranking prediction accuracy. Tokenizer versions that scored better on these metrics also had better CTR prediction accuracy.

The practical implication of this finding is iteration speed. If token quality can be assessed independently of a full model training cycle, the team can quickly screen and improve the tokenizer without waiting for complete retraining—turning a serial workflow of 'change tokens → wait for model → check results' into a parallel one of 'change tokens → check intrinsic metrics → confirm direction → then retrain.'

Meta describes this mechanism as a tokenization flywheel: tail ads gain more accurate ranking through shared content structure, the training data they produce then improves prediction quality for all ads, forming a self-reinforcing loop.

It should be noted that the 0.35% conversion lift is Meta's self-reported internal data, not independently reproduced; the eight-version correlation finding likewise comes from internal testing. The post does not disclose the token vocabulary size, codebook dimensions, or inference latency figures.

On October 19, 2026, OpenAI published a video interview titled "Thomson Reuters on Building Trust in Legal AI." In the episode, Camillo Speroni, representing Thomson Reuters, discusses the core challenge of introducing frontier artificial intelligence into the legal profession: trust. According to the official episode description, Speroni argues that for legal professionals, building trust hinges on AI-generated outputs being verifiable, citable, and suitable for lawyers to sign off on.

The content focuses on the collaboration between Thomson Reuters and OpenAI, specifically involving the integration of two legal tech products, HighQ and CoCounsel Legal, into ChatGPT Enterprise. This initiative aims to address the pain point of switching between multiple tools for legal professionals, attempting to provide the research and collaboration tools they rely on daily within a unified platform.

As a resource selection, this entry offers a specific case perspective for readers interested in enterprise AI adoption, particularly in high-compliance industries like law, focusing on overcoming barriers related to accuracy and accountability. Although a full transcript was not retrieved, the publisher's provided topic clearly points to "verifiability" and "workflow integration" as key issues in legal AI applications.

Wellington Management has released a video episode featuring Christina Kopec Rooney, the firm's Head of US Wealth, alongside Sandhya Douglas, Head of Strategic Partnerships. The program focuses on rethinking investment risk, specifically exploring how Ms. Rooney's perspective on risk has evolved over time.

As an executive at a major asset management institution, Ms. Rooney discusses the strategic selection of risks taken in pursuit of investment returns. This topic offers valuable insight into how professional institutional investors balance risk and reward in complex market environments.

Published on October 19, 2026, the episode provides an opportunity for investors to understand the internal risk management perspectives at Wellington Management. While specific arguments are not detailed in the public description, the speakers' positions and the topic's relevance make it a noteworthy resource.

Voice AI company Deepgram processes quadrillions of audio tokens a year for more than 200,000 developers and 1,400 organizations. Its offline batch workloads—contact-center analytics, medical transcription and similar scenarios—are extremely sensitive to inference cost. In October 2026, Deepgram and the AWS Forward Deployed Engineering team published a complete engineering record of porting the Nova-3 speech-to-text model to Trainium2 chips, self-reporting that a single EC2 trn2.3xlarge instance processed 11–70% more audio per dollar.

The port was not a simple compilation swap. The team first ran Nova-3 in eager mode on Trainium2 via TorchNeuron and checked which operations fell back to the CPU, then compiled with torch.compile, producing separate static-shape graphs for the encoder and decoder. The compiled path ran every model operation on Trainium2 with no CPU fallbacks detected. Numerical accuracy was validated at two levels—layer-level output consistency against CPU and end-to-end Word Error Rate—ensuring alignment with Deepgram's reference values.

The real gains came from kernel-level optimization. The team adopted a flywheel strategy: enable an NKILib pre-built kernel, profile with Neuron Explorer, then tune or hand-write a replacement based on profiling insights. Three key findings each addressed a specific bottleneck. The reference kernel for the feed-forward network layer reloaded the full set of projection weights from off-chip memory for every token tile, making memory bandwidth the limiting factor as token count grew; the hand-tuned kernel loads each weight tile once and reuses it across multiple token tiles, achieving 2.0× lower latency at batch size 96. The experimental NKILib kernel for depthwise 1D convolution framed the operation as an implicit GEMM, leaving Tensor Engine utilization low; the team rewrote it as a broadcast multiply-accumulate loop on the Vector Engine, reaching full engine utilization for a 1.5× speedup at batch size 96. The conv2d layers hit a compilation bottleneck beyond batch size 32, preventing single-graph compilation; since NKILib had no conv2d kernel, the team wrote one from scratch in NKI, unlocking batch sizes above 32 and achieving up to 50× speedup on the convolution module at batch sizes ≤32.

The final source of gain was the runtime language. Nova-3 performs a sequence of small decoder calls for each audio chunk, making host-side dispatch the bottleneck at low batch sizes. Deepgram's production serving stack is itself written in Rust, so the team built a lightweight Rust runtime: Rust coordinates eight model replicas through a shared work-stealing queue, while a C layer loads and executes ahead-of-time-compiled NEFF files directly through the Neuron Runtime API, entirely bypassing the Python interpreter. Compared with a multiprocessing Python baseline, the Rust runtime eliminates interpreter overhead; compared with a multithreaded baseline, it further alleviates GIL serialization. At the throughput-maximizing batch size, the Rust runtime delivered a 7.5% throughput gain over the multiprocessing Python baseline.

It should be emphasized that all the above figures are self-reported measurements by Deepgram and AWS. The baseline is the compute options Deepgram had previously been running, whose specific configurations are not disclosed. The wide 11–70% range covers different workloads and batch sizes; readers evaluating their own scenarios should not directly extrapolate from this range.

In the 2010 letter to shareholders, Warren Buffett detailed Berkshire Hathaway's performance measurement standards and major capital decisions. Facing market volatility, he did not pursue short-term stock price performance but emphasized establishing a mechanism for long-term intrinsic value growth. This letter not only recorded the financial results of that year but also revealed the management philosophy behind them.

First, Buffett clarified the benchmark for performance evaluation. He stated: "To eliminate subjectivity, we therefore use an understated proxy for intrinsic-value – book value – when measuring our performance." This approach aims to eliminate biases caused by subjective judgment. Although the market value of some businesses is far higher than their book value, using conservative book value as a proxy indicator can more objectively reflect whether management has truly created returns for shareholders that exceed the index.

Second, the letter focused on analyzing the logic behind the acquisition of Burlington Northern Santa Fe (BNSF) railroad company. Buffett believed that railroads have significant cost and environmental advantages over trucking: "That’s three times more fuel-efficient than trucking is, which means our railroad owns an important advantage in operating costs." This structural advantage made BNSF a core asset for enhancing Berkshire's "normal" earnings power, with pre-tax income expected to increase by nearly 40%.

Additionally, Buffett emphasized managing uncertainty from a long-term perspective. He mentioned: "Money will always flow toward opportunity, and there is an abundance of that in America." Even during economic downturns or market chaos, Berkshire continued to make large-scale capital expenditures within the United States. This persistent pursuit of opportunities, rather than reactions to macro noise, constitutes the core characteristic of its capital allocation.

For contemporary readers, understanding this mechanism helps reflect on performance attribution issues in investing. For example, if a fund manager claims their strategy outperforms the market, but their performance relies on non-replicable market timing rather than verifiable improvements in operational efficiency, sustainability is difficult to ensure. Buffett's method suggests that we should focus on decisions that can translate into stable cash flows through specific operational advantages (such as cost structures), rather than merely looking at final investment return figures.

The Berkshire Hathaway Shareholder Letter 2010 was written by Warren Buffett and published in 2010. It is suitable for investors who wish to deeply understand performance measurement standards in value investing and the logic behind large M&A decisions. It is recommended to start reading from the "Performance" section, combined with the BNSF acquisition case study.