Workslop, LinkedIn data, and autonomous retail
Plain talk on quality, policy, and shifting capabilities.
Welcome to FullStack HR, and an extra welcome to the 87 people who have signed up since last edition.
If you haven’t yet subscribed, join the 9800+ smart, curious and like-minded future of work people by subscribing here:
Happy Friday,
I think THE most common question I get when talking about AI is “How do I create a PowerPoint presentation with my template?” And the answer has been that it’s not really possible (Gamma has been the best solution so far) but a couple of weeks ago, Anthropic introduced the capability into Claude! Watch the video below and I’ll show you how!
And if you speak Swedish, I’ll do a free update on Monday on all things AI. 700+ people have already signed up, and I’d love to see you there!
Now, let’s get to it!
🧠 Topics I’m engaging with
AI-generated “workslop” is destroying productivity HBR + BetterUp Labs coin “workslop” for polished-looking, low-value AI output. The pattern is familiar: leaders mandate tools without training, norms, or evaluation, then express shock when quality dips. The fix is operational: explicit standards, enablement for managers, and review loops that reward decision quality over output volume.
My response on LinkedInLinkedIn will train AI on member profiles . Starting Nov 3, 2025, certain profile fields, posts, and public activity may be used for model training (under-18s excluded). It’s on by default, and to opt out, go to Settings → Data privacy.
Anthropic Economic Index (Sept 2025). This report examines the economic impact and adoption of AI technologies, focusing on disparities between different geographic regions and types of enterprises. Usage is uneven by region and industry, and enterprise API traffic is majority automation-oriented. The real constraint isn’t model price but integration quality and decision architecture. Design around measurable outcomes and guardrails, not just tokens and prompts.
Fiverr lays off 30% in an AI push Fiverr leans further into workflow automation. Expect “describe-and-deliver” tasks to commoditize; the defensible edge shifts to domain context, creative direction, and durable client relationships.
Google developer survey: no “AI overload” (yet) High adoption with pragmatic use: AI handles routine coding and reviews; humans focus on problem framing and architecture. For non-dev teams, mirror this: automate the grunt work, keep judgment with humans, measure impact at the workflow level.
SHRM Executive Network — HR Tech Trends (Sept 2025) Monthly CHRO digest worth bookmarking: vendor shifts, AI governance maturity, and workforce design themes. Use as a checklist for your Q4/board update—policy clarity, risk controls, and a handful of credible ROI cases.
⚙️ Tools to try
Bettabite - Bettabite is a skills-first hiring platform that uses visual charts to match people and teams based on skills and work style, not just resumes. It highlights fit, gaps, and growth potential to speed up hiring decisions
Anthropos - Anthropos lets companies create custom AI-driven simulations for team training, screening, onboarding, and skills assessment. Scenarios are fully customizable and available in multiple languages, with online delivery
ElevenLabs Agents - Fast path to voice/multimodal assistants for HR (manager coaching, policy Q&A, intake). Strong at guided conversations, tone control, and handoff to humans!
🔁 Important updates
ChatGPT Pulse is a new OpenAI feature (in preview) that gives Pro users daily, personalized updates and insights, with optional Gmail and Calendar integration. It brings proactive, tailored support directly to your device.
AI system that writes expert-level research code - Preprint shows LLM + search generating novel methods that beat human baselines; value shifts from coding to problem framing and evaluation.
Xpand opens a fully autonomous 24/7 store in central Vienna - Vision + robotics enable staff-free retail; expect new role mixes (remote ops, field tech) and tighter privacy/safety governance.
“Pick manual jobs—AI is taking office work” (viral thread) - Spiky take on a graduate-jobs crunch; real action is redesigning early-career ladders with AI co-pilots and real supervision.