Loading…
Latest reads
Every post, summarised
Grounded summaries of engineering blogs, newest first. Each links to the original.
- 4,671 posts
- 20 companies
The Good Stuff
Paid picks we use ourselves. Links carry a referral code — a subscription through one earns us a commission at no extra cost to you.
Latest posts
Higher usage limits and a SpaceX compute deal
An agreement with SpaceX will give the company access to all compute capacity at the Colossus 1 data center, adding more than 300 megawatts and over 220,000 NVIDIA GPUs within the month. Combined with other compute partnerships, the expansion supports higher usage limits: Claude Code five-hour limits double for Pro, Max, Team, and seat-based Enterprise plans, peak-hour reductions end for Pro and Max, and Claude Opus API limits rise considerably. The company says the new capacity will directly improve service for Claude Pro and Claude Max subscribers while it trains and runs Claude on AWS Trainium, Google TPUs, and NVIDIA GPUs. It is also expanding inference capacity in Asia and Europe and exploring orbital AI compute with SpaceX.
Dario Amodei's remarks at the AI Safety Summit
At the AI Safety Summit, Dario Amodei explains why rapid AI progress makes dangerous capabilities difficult to predict and motivates Anthropic’s Responsible Scaling Policy (RSP). The policy combines AI safety levels (ASLs), modeled loosely on biological safety levels, with recurring tests for dangerous capabilities along the compute scaling curve. ASL-2 covers current models and calls for model cards, external red-teaming, and strong security, while ASL-3 would require unusually strong protection against weight theft and safeguards preventing operationally useful CBRN information; ASL-4 adds escalated misuse and autonomous-system risks. Amodei also describes executive involvement, integrating RSP deadlines into product and research planning, board and external accountability, and the goal of refining company policies into regulatory testing and auditing regimes.
Expanding access to safer AI with Amazon
Anthropic announces that Amazon will invest up to $4 billion and become its primary cloud provider for mission-critical workloads. The collaboration combines Anthropic’s frontier safety research and models with AWS infrastructure, including Trainium and Inferentia chips, while expanding Amazon Bedrock support for secure model customization and fine-tuning. Organizations will be able to use Claude 2 through Bedrock, including its 100,000-token context window, with examples spanning legal search, investment analysis, and travel recommendations; Lonely Planet reduced itinerary-generation costs by almost 80 percent after deployment. Amazon will take a minority stake, while Anthropic’s governance remains unchanged under the Long Term Benefit Trust and its Responsible Scaling Policy, including pre-deployment testing of new models.
Prompting Claude's long context window
An evaluation of Claude’s 100,000-token context window tests prompting techniques for recalling specific information from long documents. Researchers built randomized collages from sections of a July 13 government document, using Claude-generated multiple-choice questions and filtering out questions it could not answer from the source passage alone. They compared a baseline with unrelated examples, dynamically selected contextual examples, and two or five examples, each with or without a scratchpad that instructed Claude Instant 1.2 to extract relevant quotes across roughly 70K and 95K token contexts. Scratchpads and contextual examples generally improved performance, with more examples performing better, although recall could decline when the answer passage was at the end. For Claude 2, the best prompting strategy raised performance from 0.939 to 0.961, a reported 36% reduction in errors; reproducible code was released in the Anthropic Cookbook.
Anthropic partners with BCG
Anthropic announces a collaboration with Boston Consulting Group (BCG) to give BCG customers around the world direct access to Claude and support enterprise AI deployments. BCG will advise customers on strategic AI applications and help deploy Anthropic models, including Claude 2, across use cases such as knowledge management, market research, fraud detection, demand forecasting, report generation, and business analysis. The partnership is grounded in responsible AI, with Anthropic citing Constitutional AI and a shared focus on helpful, honest, harmless, safer, and more reliable systems. BCG has also used Claude within its own teams and previously worked with Anthropic, including at the United Nations, while the companies say their goal is to align business impact with ethical deployment.
Releasing Claude Instant 1.2
Claude Instant 1.2 is available to businesses through an API as a faster, lower-priced model for casual dialogue, text analysis, summarization, and document comprehension. It incorporates strengths of Claude 2 in real-world use cases, producing longer, more structured responses, following formatting instructions better, and improving math, coding, reasoning, safety, quote extraction, multilingual capabilities, and question answering. Against Claude Instant 1.1, it scored 58.7% versus 52.8% on the Codex evaluation and 86.7% versus 80.9% on GSM8K, indicating gains on the reported math and coding measures. The release also reports less hallucination and greater resistance to jailbreaks in automated red-teaming, while developers can call the model through the API; pricing for developers is available separately.
Claude 2
Claude 2 is introduced as a new model with improved performance, longer responses, expanded coding, math, and reasoning capabilities, and access through both an API and the claude.ai beta website. It supports prompts of up to 100K tokens and can produce documents of a few thousand tokens, enabling work across hundreds of pages of technical documentation or a book. Reported results include 76.5% on the Bar exam multiple-choice section versus 73.0% for Claude 1.3, 71.2% on Codex HumanEval versus 56.0%, and 88.0% on GSM8k versus 85.2%. Anthropic says internal red-team evaluation found Claude 2 2x better at giving harmless responses than Claude 1.3, while acknowledging that jailbreaks remain possible and inappropriate outputs can occur. The API is offered to businesses at the same price as Claude 1.3, and the chat experience is available in the US and UK, with broader availability planned.
Anthropic raises $450 million in Series C
Anthropic announces a $450 million Series C funding round led by Spark Capital, with participation from Google, Salesforce Ventures, Sound Ventures, Zoom Ventures, and others. The funding is intended to support development of helpful, harmless, and honest AI systems, including Claude, and to expand product offerings, support for businesses deploying Claude, AI safety research, and the company’s product and research teams. Anthropic says its alignment work targets adversarial conversations, precise instruction following, and greater transparency about system behavior and limitations. Planned product work includes 100K context windows for submitting hundreds of pages of material to Claude, along with further improvements to its API and chat products. Yasmin Razavi, a Spark Capital general partner, joined Anthropic’s board, while the company continues refining its governance structure around its Public Benefit orientation.
Partnering with Scale to bring generative AI to enterprises
Anthropic announces a partnership with Scale to bring Claude and generative AI applications to enterprises. Scale customers can now use Claude, Anthropic’s conversational AI assistant. Businesses can build applications on Claude using Scale’s deployment and management functionality, with expert prompt engineering and model validation available to improve performance and identify weaknesses. Scale’s enterprise-grade security is available through a customer’s private AWS environment, while data connectors can import proprietary sources including databases, Confluence, Google Drive, and Outlook. The post describes the combination of Claude, Anthropic’s Constitutional AI system, and Scale’s tooling as an enterprise-ready solution, and says the partners aim to bring Claude to more organizations responsibly.
An AI policy tool for today: Invest in NIST
Effective AI policy depends on accurately describing and quantifying systems’ capabilities and risks, but widely agreed methods for comprehensive assessment remain lacking. The proposal calls for ambitious federal funding for NIST to expand its AI measurement and standards work, building on efforts including the Face Recognition Vendor Test and AI Risk Management Framework. Additional resources could support standardized measurement techniques and community infrastructure such as testbeds for open-ended systems, with Anthropic recommending a $15 million increase over FY 2023, inclusive of NIST’s requested $5 million increase. The post presents this as a “shovel-ready” complement to internal controls, independent audits, and public-interest regulation. It argues that rigorous evaluation could identify risks before deployment, improve public trust and governmental confidence, and create incentives for safer innovation.
Introducing Claude
Anthropic is offering Claude more broadly after a closed alpha with partners including Notion, Quora, and DuckDuckGo, positioning it as an AI assistant for conversational and text-processing work. Available through a chat interface and API in the developer console, Claude supports summarization, search, writing, question answering, coding, and other tasks, while accepting direction on personality, tone, and behavior. The announcement introduces Claude, a high-performance model, alongside Claude Instant, a lighter, less expensive, faster option, and says both are being developed around helpfulness, honesty, harmlessness, reliability, and predictability. Partner deployments include Poe, Juni Learning’s Discord tutor, Notion AI, DuckAssist, Robin AI’s contract tools, and AssemblyAI’s audio platform. Reported benefits include natural conversation, richer tutoring responses, contract drafting and explanation, improved engagement, and stronger feedback, though these are partner and user reports rather than a controlled evaluation.
From One Horizontal Scaling Controller to Many: Evolving Uber's Compute Platform
Uber’s failover redesign exposed a limitation in its long-standing single-controller model for scaling stateless workloads. To reuse idle capacity during regional failover, the platform needed multiple orchestrators to influence replica counts without overloading UDC, which remained on the normal deployment lifecycle path. The solution introduced the ServiceScale CRD and Service Scale Controller, allowing separate actors to record steady-state and temporary failover intent in Kubernetes and letting SSC reconcile the combined decision. Production work then exposed risks from informer caches that lagged reality and near-simultaneous multi-writer updates, including ReplicaSet metadata-spec drift that disrupted proportional scaling and could leave workloads stuck. Uber added fleet-wide detection, an automated UDC healer, a long-term scaling-path fix, and extensive kind integration tests, staging, and canaries; the year-long rollout avoided customer-impacting outages while establishing a more extensible control plane.
Anthropic raises Series B to build robust AI
Anthropic, an AI safety and research company, raised $580 million in Series B financing to build large-scale experimental infrastructure for studying and improving the safety properties of computationally intensive AI models. Since its 2021 founding, the company has researched steerability, robustness, and interpretability, including mathematically reverse engineering small language models and investigating pattern-matching behavior in larger ones. It developed baseline techniques and reinforcement learning methods for making models more “helpful and harmless,” released an alignment-focused dataset, and analyzed sudden performance changes and their societal impacts. The stated goal is to create models with stronger implicit safeguards, fewer after-training interventions, and tools for examining whether those safeguards work, alongside policy and societal research. The round was led by FTX CEO Sam Bankman-Fried and followed a $124 million Series A in 2021; Anthropic said its team numbered around 40 people.
Anthropic raises $124 million Series A
Anthropic, an AI safety and research company led by CEO Dario Amodei and President Daniela Amodei, has raised $124 million in a Series A financing round. The funding will support its research roadmap, computationally intensive work, and prototypes of reliable and steerable AI systems, with an emphasis on systems that are interpretable and robust. Its immediate priorities include increasing the reliability of large-scale AI models, developing techniques and tools for interpretability, and integrating human feedback more tightly into AI development and deployment. The round was led by Jaan Tallinn, co-founder of Skype, and included James McClave, Dustin Moskovitz, the Center for Emerging Risk Research, Eric Schmidt, and others; Anthropic is hiring researchers, engineers, and operational experts.
Five AI Questions We're Hearing from Financial Services Leaders
Financial services leaders are shifting from asking whether AI works to asking whether it can earn trust, amid governance and audit concerns. The post frames five operational questions around moving finance from reporting to action, using AI in financial crime without losing control, supporting bankers and investment teams, using real-time data, and managing expanding costs. Examples include governed GL-to-Risk reconciliation with cited evidence, human maker-checker oversight and an audit trail; AML agents that draft SARs while humans file; and multi-agent workflows for client and portfolio analysis. The post also cites Coinbase's sub-100ms P99 fraud-detection benchmark and warns that token costs manageable in pilots may grow in production, making model routing and spend visibility relevant. Databricks presents these topics through live demos, talks and meetings at Sibos 2026 in Miami, scheduled for September 28–October 1, including a session on governing AI in financial-crime operations.
Announcing ClickHouse Managed Postgres on Google Cloud
ClickHouse Managed Postgres is entering Private Preview on Google Cloud, extending the fully managed, NVMe-backed Postgres service previously launched on AWS. The GCP deployment targets teams wanting Postgres for OLTP and ClickHouse for OLAP in one cloud, avoiding cross-cloud replication, egress costs, and separate networking and compliance management. It uses GCP Local SSD storage, native change data capture through ClickPipes, in-region networking, and pg_clickhouse, which pushes filters, joins, and aggregations to ClickHouse behind a single Postgres connection. Preview capabilities include zonal high availability with two standbys and quorum replication, point-in-time recovery, read replicas, PgBouncer, more than 90 Postgres extensions, monitoring, Query Insights, scheduled upgrades, and infrastructure-as-code support through OpenAPI, Terraform, and clickhousectl. Access is invitation-only via a waitlist, with availability in every GCP region where ClickHouse Cloud runs and hands-on migration support from Cloud SQL, AlloyDB, RDS, Aurora, Neon, or self-managed Postgres.
Why Rider and ReSharper Were Slow to Start, and How Microsoft Helped Fix the Problem
After ReSharper’s out-of-process (OOP) architecture launched, users reported slower Windows startup, and profiling traced the delay to Microsoft Defender scanning the separate process. Defender spends only a couple of seconds on processes in write-protected paths but can add tens of seconds when ReSharper loads DLLs from a user installation directory. The investigation used Microsoft Defender ETW logs and direct CPU-time measurements, with cold-start tests repeated 10 times inside an 8-vCPU, 8-GB Hyper-V VM; JetBrains IDEs took roughly 10–40 seconds to scan, while Microsoft IDEs and other editors stayed under one second. Collaboration with Microsoft led to Defender optimizations in version 1.449.454.0, improving scanning for ReSharper OOP and Rider in write-protected directories. JetBrains also released the Defender Performance Tool for live monitoring, offline snapshots, and CSV export.
Announcing ‘The 2026 AI Sentiment Report’: How end users feel about AI Agents
The 2026 AI Sentiment Report examines how end users feel about AI Agents in customer service and other uses such as shopping. Overall sentiment improved: 49% reported positive experiences, up nine percentage points from 2025, and that figure rose to 74% after respondents watched an Agent resolve a real query. Trust remains conditional, with 54% willing to trust AI with simple or routine issues but not complex ones involving discretion, privacy, continuity, accountability, or judgment. The research identifies easy escalation to a human and upfront disclosure that users are interacting with AI as the two leading ways to increase comfort, while trust also varies by channel, with chat strongest and social weakest. It concludes that businesses can treat trust as a design problem, including reassuring customers that shopping assistants serve their interests rather than merely the company’s financial goals.
A practical approach to end-to-end Solvency II reporting in Databricks
Solvency II reporting connects data, actuarial reserving, capital calculation, quantitative reporting templates (QRTs), the ORSA, approvals, and disclosure, but fragmented systems make readiness and cross-process questions difficult to manage. The described Databricks implementation provides a governed workflow and control tower around existing actuarial and capital suites, automating ingestion checks, data-quality review, reconciliation, orchestration, reporting, and audit trails. It uses MLflow-managed reserving models alongside engines such as Prophet, RAFM, and Igloo, while narrow-scope AI agents can investigate QRT mismatches, suggest remediation, route questions, and support human review. An LLM can draft an ORSA from current numbers, while scenario analysis assesses doubling a cyber book over 12 months; reviewers retain decision authority. The workflow is presented as easier to monitor and investigate.
PwC deploys Claude across its business
Anthropic and PwC are expanding their strategic alliance to use Claude for technology delivery, deal execution, and AI-led redesign of enterprise functions. PwC will roll out Claude Code and Cowork from U.S. teams toward a global workforce of hundreds of thousands, establish a joint Center of Excellence, and train and certify 30,000 professionals. The collaboration targets agentic technology builds, AI-native deal-making, and production operating models for finance, supply chain, HR, and engineering, alongside a new Claude-based Office of the CFO for regulated industries. Reported deployments include insurance underwriting reduced from ten weeks to ten days, an HR application delivered in under two months, and cybersecurity response accelerated from hours to minutes, with improvements of up to 70%.
Top topics
- Architecture834
- Performance816
- LLMs660
- Machine Learning622
- Open Source563
- AI Agents553
- AI470
- Developer Experience448
Weekly digest
The best engineering reads, once a week.