Tag
The post argues that AI agent evaluations should measure total workflow effort—including human verification, corrections, and failure costs—rather than just task completion speed, since review overhead can outweigh time saved.
The article argues there is no single AI adoption rate, comparing survey-based reach, transaction-based paid demand, production deployment, and financial-return measures from sources like McKinsey, Ramp, and the US Census, and explains why their differing denominators, visibility, and time periods make the figures non-comparable.
A practitioner outlines the top 7 reasons enterprise AI pilots fail — starting with no process map, automating an immaterial step, and lacking a baseline — arguing process re-engineering must precede AI deployment.
一年后回顾MIT"95%的AI试点失败"这一广为引用的数据：该数字本身是基于对报告漏斗的误读和仅153份问卷的垃圾统计，但后续来自McKinsey、Deloitte、Capgemini及政府统计的数据表明，其背后的诊断——多数企业AI支出仍未反映到损益表上——大体成立。
作者重新计算 AI 客服 agent 的 ROI：原以为 78% 自动关单、11% 重开率很好，但按客户+订单号+意图标签匹配 7 天内的新工单后，发现约 13% 的"已解决"其实是同一问题重开，实际近四分之一仍需人工，单次成功任务成本从 $0.54 升到 $0.71。文中用 Pointfive 做调用追踪、Zendesk 存工单，并询问重复工单的最佳匹配键。
A practitioner shares the top 25 questions large-enterprise C-suite executives ask about AI transformation — from workflow selection and agent sprawl to ROI, data governance, and headcount implications — as a lead-gen post for their firm Varick.
NVIDIA argues its AI factories maximize return on investment by being productive (highest tokens per megawatt), durable (GPUs like the A100 keep earning for years), and fungible (able to run any workload), citing SemiAnalysis data showing Vera Rubin NVL72 delivering over 30x higher throughput per megawatt than GB300 NVL72.
The tweet speculates on the potential financial impact for an engineer if NVIDIA's $12.9 billion acquisition of Hugging Face goes through, emphasizing a high return on equity despite dilution.
An article exploring the anticipated developments in AI by 2026, with an emphasis on realizing return on investment.
The article argues that AI adoption in enterprises often fails because companies focus on buying tools rather than redesigning workflows, leading to low ROI and minimal impact. It emphasizes the need for architectural changes to achieve real success.
The article reports that business adoption of AI agents has tripled this year, with companies moving from testing to scaling deployments and achieving measurable ROI, based on Salesforce's Agentic Enterprise Index.
Navigara is a product that aims to provide measurable return on investment for AI coding expenditures.
KPMG survey finds 49% of large organizations have narrowed, delayed, or paused AI agent deployments as operating costs exceed value, with only 7% achieving established ROI despite rising adoption and confidence.
A new field called 'tokenomics' has emerged to measure the return on investment for the massive sums companies are spending on artificial intelligence.
A large Danish study finds that AI saves workers about 2.8% of their time, but these gains don't translate into measurable business value because organizations fail to intentionally reallocate the freed capacity.
Google's signal of larger AI infrastructure spending shifts the debate from whether companies will invest in AI to which companies will profit from it, as investors focus on ROI.
A discussion thread asking enterprise professionals whether building custom in-house agent memory (RAG, knowledge graphs, etc.) is worth the investment compared to relying on built-in cloud LLM memory features, exploring ROI, data risk, and complexity.
Enterprise finance teams are struggling with AI implementations, but dedicated background agents can automate repetitive tasks like invoice matching and bank reconciliation, delivering measurable ROI. Varick Agents claims to have helped clients reduce month-end close from 12 to 5 days and save $45M annually.
An analysis of AI agent adoption in 2026, citing Gartner, McKinsey, and Deloitte data showing that while over 40% of agentic AI projects are forecast to be cancelled by 2027 and fewer than 10% of enterprises have scaled one for real value, coding agents have become default practice with over 90% deployment. The article finds the predicted 'wow' use case is uneven, concentrated in coding assistance and document retrieval, while governance remains immature.
An analysis piece discussing that the return on investment for AI outside the tech sector may take longer to materialize, as non-tech industries face adoption challenges.