AI at Work
AI adoption inside companies and jobs: enterprise rollouts, AI-native operations, productivity, labor-market effects, AI in specific industries.
Merantix Momentum advises European firms on AI sovereignty
Berlin-based consultancy Merantix Momentum is helping European clients navigate AI sovereignty and business viability. The firm argues that AI strategy must now be inseparable from core business strategy.
Measuring AI impact requires tracking engine metrics over financials
Merantix Momentum advises that direct monetary ROI from AI is often delayed and rare. Organizations should instead track leading indicators like process cycle time and quality.
Sharplink ceo projects ai agents will save consumers $1.4 trillion by 2035
Joseph Shalom estimates that AI agents will intermediate over 35% of financial services fees by 2035. He argues that this shift will optimize portfolios and move idle cash to higher-yield opportunities, provided regulatory guardrails are established.
Gates warns ai will cheaply replace white-collar labor in key sectors
Bill Gates asserts that current AI exceeds human cognition in specific tasks, making white-collar labor replacement inexpensive. He argues that existing tax policies accelerate this displacement and calls for new social safety nets.
AI agents allow small teams to run high-gmv marketplaces
Personal agent platforms are reducing the operational overhead required to manage digital marketplaces. This shift enables teams of five or fewer to support businesses generating millions in gross merchandise value.
Emerging agent-to-agent communication creates new social and serendipitous connections
Developers are exploring how AI agents can facilitate social interactions and serendipity beyond simple task completion. This paradigm shift aims to use agents as social lubricants that connect users based on shared interests.
Musk cites unsustainable costs for GrokBot, sparking debate on agent viability
Elon Musk stated that serving GrokBot is cost-prohibitive relative to revenue, contradicting prior claims of high growth. The admission has intensified debate over whether personal AI agents can be monetized at current price points.
Upwork deploys internal AI tool to democratize business insights
Upwork CEO Hayden Brown described an internal tool called Crystal Ball that provides governed access to structured business data. The tool allows individual employees to obtain insights in hours that previously required weeks of analyst work.
WatchDuty CEO emphasizes human oversight in AI-driven crisis response
WatchDuty CEO John Mills argued that human interpretation is essential in AI-driven crisis response to avoid fatal errors. He described his company as a signals intelligence organization where AI augments, rather than replaces, human judgment.
Students advocate for earlier AI integration in German school curricula
German students argue that AI education begins too late in the school system, often only in 10th grade. They propose starting instruction in 5th grade to align with the age at which students first receive smartphones.
Major banks block AI agents from managing customer accounts
JP Morgan, Bank of America, Citi, and Schwab have stated they will not allow outside AI agents to move money on behalf of customers. Analysts argue this restriction protects bank moats but represents a temporary barrier to irreversible technological progress.
Lagora automates legal due diligence with multi-model AI agents
Stockholm-based Lagora has deployed an AI agent platform that processes up to one million cells of legal data simultaneously, aiming to shift law firm billing from hourly rates to project fees. The company’s multi-model strategy routes tasks to specific LLMs based on cost and capability, with recent updates improving citation accuracy from 70% to 99%.
Study finds 40% of US desk workers receive AI-generated slop
A joint study by Stanford and BetterUp found that 40% of U.S. desk workers received low-quality AI-generated work in the past month, costing recipients nearly two hours. Experts propose frameworks to distinguish high-quality AI assistance from generic output, emphasizing human control over narrative architecture.
Modal spends $500,000 monthly on coding agents for engineers
Modal allocates approximately $500,000 per month to coding agents for its 150 engineers, arguing that AI increases software demand rather than eliminating jobs. The company is competing for talent against AI labs offering packages of $20 million or more.
AMP launches AI agent Ray for renewable energy monitoring
AMP is deploying an AI agent named Ray to monitor renewable energy portfolios in Sub-Saharan Africa, using code execution to triage alerts and quantify financial impact. The company has also standardized its engineering workflow on Slack-integrated agents to improve auditability and collaboration.
Bending Spoons reports 94% of code is now AI-written
Bending Spoons CEO Luca Ferrari states that 94% of the company's code is now AI-generated. He argues that no good company should be below 70% AI-written code at this stage.
AI agents form human-like relationships through tool use
Sahil Jain argues that combining models with tools enables AI agents to form relationships with humans. He believes this trend will fundamentally change work and knowledge economies.
Marketers should keep final creative assets human-made
Hector Serrano argues that using AI for final customer-facing content risks damaging brand perception. He advises reserving AI for internal research while humans handle final production to maintain quality.
Venkatesh Rao links AI panic to cognitive mercantilism
Venkatesh Rao argues that current fears about AI stem from an outdated view of intelligence as a scarce resource. He predicts AI will expand demand for cognitive labor rather than eliminate it.
AI demand elasticity challenges fixed labor assumptions
NLW argues that the lump of labor fallacy ignores the elasticity of human demand. He contends AI will expand total work by lowering barriers across six specific dimensions.
Arca agents detect hidden fees and non-fiduciary advisors
Arca’s AI agents analyze client documents to identify financial issues often missed by humans, including unauthorized fees and compliance risks. This capability allows the firm to offer a level of continuous monitoring previously impossible in traditional advisory models.
DoorDash reports minimal order volume from third-party AI agents
DoorDash disclosed that third-party AI agents have generated only several hundred orders, a negligible fraction of its 3 million daily deliveries. CEO Tony Hsu described the sector as dynamic but cautioned against overestimating immediate disruption.
Hastings predicts AI tutors will surpass human teachers in four years
Reed Hastings predicts AI will outperform typical teachers within two to four years, offering affordable one-to-one instruction. He argues this technology will make personalized tutoring, currently a luxury, accessible to all students.
AI models cut drug discovery time and cost for Citrini
Citrini Therapeutics used Anthropic models to identify molecules for scorpion stings in 48 hours, bypassing traditional antivenom methods. The project cost $200,000 and was completed by a team with no biotech background.
OpenAI launches ChatGPT for Financial Services with data integrations
OpenAI released a financial services product built with Morgan Stanley and Evercore, integrating data sources like Dilupa and S&P. The company stated that third-party connectors had high error rates, necessitating in-house data integration.
Crosby bids to acquire bankrupt law firm Weil's data assets
Legal AI startup Crosby has submitted a letter of intent to buy the data assets of Weil, Gottschall & Manjiz, which recently filed for Chapter 11 bankruptcy. The move aims to secure a massive archive of legal precedents for AI training, though critics argue the client data is legally non-transferable.
Non-engineers build personal finance and health apps using AI agents
A new class of non-engineers is using AI agents to create personal applications for finance, health, and productivity. This trend shifts software creation from enterprise-focused tools to individualized, agent-managed personal infrastructure where users own their data.
Anthropic study shows experts often adopt AI agent recommendations
An experiment by Anthropic found that 86% of statistics experts chose an AI agent's pre-selected option, even when they had previously decided otherwise. The findings raise concerns about scientific integrity and the independent verification of results when using the same models.
Eve shifts legal software pricing from seat-based to outcome-oriented models
Eve, an AI legal platform, is moving away from traditional seat-based SaaS pricing to outcome-oriented models that align with contingency fees. The platform allows plaintiff firms to manage five times more clients by automating labor-intensive tasks.
Evaluation frameworks become critical for production AI agent reliability
As companies move AI agents into production for customer-facing roles, evaluation harnesses are becoming essential to ensure reliability and policy adherence. Metrics have expanded to include task completion, cost, latency, and safety, determining go/no-go decisions for releases.
Bayer deploys agentic AI assistants Friday and Eli to thousands of users
Bayer is piloting an internal agentic AI system named Friday for 1,000 employees and has scaled its sales assistant Eli to 1,200 users across eight countries. The deployments focus on non-mission-critical tasks and sales support, with plans for broader rollout.
Cisco deploys personal AI agents to 90,000 employees
Cisco has rolled out personal AI agents to its workforce, with token usage reaching two trillion in a single week. The agents automate tasks from meeting transcription to dashboard generation.
Microsoft offers voluntary buyouts to 60,000 employees
Microsoft is offering buyouts to 27% of its workforce to manage costs as AI infrastructure bills come due. The move follows a previous round of layoffs and mirrors trends at other tech firms.