• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    company
    The New Stack

    The New Stack

    22 stories tagged by Digg

    Recent stories

    Technology

    AI search webinar explores why reranking can’t fix bad retrieval

    The New Stack is inviting readers to a virtual event with Vespa’s Bonnie Chase and Jenny Morris and host Dr. Kim Fessel. It asks what happens when the right answer never reaches a search system’s candidate set.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    One engineer reportedly shipped 2,000 PRs a month to production

    The New Stack credits verification for that reported output and describes isolated, virtualized testing environments as a way to help AI agents ship thousands of PRs each month.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    Kev skips text generation for simple AI-agent decisions

    The New Stack describes Kev as a new open-source model family that makes choices in a single forward pass instead of generating text.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    How confidential AI could protect enterprise data and models

    The New Stack says an AI model can redact customers’ personally identifiable information, but no one can be fully sure the redaction worked.

    The New StackTN
    2 Sources, 10d ago, updated 10d ago

    Technology

    Claude Opus 5.5 vs. Fable 5.1 on cost, speed and 15 test runs

    The New Stack reports that Opus 5.5 got 13 of 15 runs right at 22% less cost. Fable 5.1 got all 15 and finished 40% faster, but took what the outlet calls a problematic shortcut to pass one test.

    The New StackTN
    3 Sources, 10d ago, updated 10d ago

    Technology

    Enterprise data protection and AI model reliability

    The New Stack shares a guide on protecting enterprise data without sacrificing AI model reliability, referential integrity or engineering velocity. It thanks Perforce.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    Grok Build and Claude Code put to a cross-repo memory test

    The New Stack reports that both coding agents were given the same rule, but only one remembered it in the next repo.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    A critical test-database flaw and the limits of severity scores

    The New Stack uses a critical-severity finding on a disposable test system to argue that a CVSS score doesn't show which flaws actually cost money.

    The New StackTN
    1 Source, 11d ago, first seen 11d ago

    Technology

    Claude Opus 5.5 versus Opus 5 on cost, speed and reasoning

    The New Stack says Anthropic claims Opus 5.5 is 40% cheaper and 30% faster than Opus 5. On hard reasoning tests, the savings held up, but the speed claim fell short.

    The New StackTN
    1 Source, 12d ago, first seen 12d ago

    Technology

    Tests show prompt injections can spread through email, files and code comments

    The New Stack reports that malicious instructions moved through those channels in tests and caused AI agents to take unauthorized actions.

    The New StackTN
    The RegisterTR
    3 Sources, 11d ago, updated 11d ago

    Technology

    The case for checking retrieval before upgrading a reranker

    The New Stack argues that reranking gets expensive quickly and cannot fix a search that failed to surface the right results in the first place.

    The New StackTN
    1 Source, 12d ago, first seen 12d ago

    Technology

    Human oversight in defining requirements for AI coding

    The New Stack argues that automated test gates can’t catch flawed specs, so people need to focus on defining requirements rather than writing code.

    The New StackTN
    1 Source, 12d ago, first seen 12d ago

    Technology

    AI agent debugging beyond error logs

    The New Stack reports that Nvidia backs SAFE, a shared failure-reporting system, and says debugging agents means tracing decisions, not just logging errors.

    The New StackTN
    1 Source, 13d ago, first seen 13d ago

    Technology

    The operational-ownership case for AWS Elastic Beanstalk

    The New Stack argues that Elastic Beanstalk gives lean development teams operational ownership and removes 3 a.m. production stress across app portfolios. Its post thanks AWS Marketplace.

    The New StackTN
    1 Source, 13d ago, first seen 13d ago

    Technology

    Secrets-safe context boundaries for AI coding agents

    The New Stack recommends securing local context boundaries to protect secrets before sensitive data leaves your environment.

    The New StackTN
    1 Source, 13d ago, first seen 13d ago

    Technology

    OpenSearch Webinar on PPL and Unified Alerting

    Live demo covers cross-signal alerts across logs, metrics, and traces.

    The New StackTN
    1 Source, 40d ago, first seen 40d ago

    Technology

    Finance Stalls Agentic AI Projects Over Telemetry Growth

    The New Stack reports rising data volumes are shifting focus to cost control in AI systems.

    The New StackTN
    1 Source, 40d ago, first seen 40d ago

    Technology

    Shai-Hulud Worm Targets Package Registries

    The New Stack reports on registry control enabling automated supply chain attacks.

    The New StackTN
    1 Source, 40d ago, first seen 40d ago

    Technology

    AI-Generated Code Breaks Traditional Code Review

    The New Stack notes engineers face diffs from models they do not control.

    The New StackTN
    1 Source, 40d ago, first seen 40d ago

    Technology

    Agentic AI Latency Arises From Iterative Processes

    The New Stack links agentic AI delays to iterative multi-agent workflows instead of hardware limits.

    The New StackTN
    1 Source, 41d ago, first seen 41d ago

    Technology

    The New Stack and Dynatrace Host Welcome Reception

    Event at The Tech Interactive in San Jose on day zero of the congress.

    The New StackTN
    1 Source, 41d ago, first seen 41d ago

    Technology

    Test Data Delays Block AI Code Generation Speed Gains

    Test data wait times slow AI adoption more than code generation does.

    The New StackTN
    1 Source, 42d ago, first seen 42d ago

    Elsewhere on the web

    @thenewstack