Erik Edin

”Look at this awesome dashboard Jane from finance made with Claude!” Finance teams are getting serious about using AI and data. Finally, they can create the dashboard that is exactly as they want it. Yes hosting, access controls and data refreshes might not work, but these are solvable problems. But before you launch a Lovable for dashboard startup, consider this: do we even need dashboards anymore? Dashboards are generally used for three things: to do basic analysis, to prepare presentations, and to track KPIs. If you give it clean data with documentation, Claude/Codex does business analysis tasks better and faster than most humans today (including myself). Claude Cowork generates the presentation PowerPoints with all the charts and analysis, no need to copy from a dashboard. Maybe we’ll keep a few dashboards to track KPIs, but I bet no one with Claude/Codex hooked up to the data will look at the dashboards. They will just ask the AI. So while finance teams are excited about finally being able to build dashboards, they might not need them for much longer. What about chat bots inside dashboard tools like Metabase and Lightdash? At the moment, Claude/Codex (with data access) does much stronger analysis. The AI models inside the dashboarding tools are also too limited in what they can do. They can’t connect to other MCPs, they don’t acreage slides, and they don’t allow you to upload your own spreadsheets. It seems like AI is simultaneously making it easier to create dashboards, but also less necessary. And maybe this applies to all vibe coding of internal tools.

There’s something interesting about how AI tools simultaneously made it easier to build software and less necessary to do so.

I think I finally get the point of MCPs.

The victory of the opposition in Hungary yesterday, like the Polish election in 2023, is a victory for democracy, not just in Europe but around the world. Most of all, it’s a testament to the resilience and determination of the Hungarian people – and a reminder to all of us to keep striving for fairness, equality and the rule of law.

Another week on the road meeting with a couple dozen IT and AI leaders from large enterprises across banking, media, retail, healthcare, consulting, tech, and sports, to discuss agents in the enterprise. Some quick takeaways: * Clear that we’re moving from chat era of AI to agents that use tools, process data, and start to execute real work in the enterprise. Complementing this, enterprises are often evolving from “let a thousand flowers bloom” approach to adoption to targeted automation efforts applied to specific areas of work and workflow. * Change management still will remain one of the biggest topics for enterprises. Most workflows aren’t setup to just drop agents directly in, and enterprises will need a ton of help to drive these efforts (both internally and from partners). One company has a head of AI in every business unit that roles up to a central team, just to keep all the functions coordinated. * Tokenmaxxing! Most companies operate with very strict OpEx budgets get locked in for the year ahead, so they’re going through very real trade-off discussions right now on how to budget for tokens. One company recently had an idea for a “shark tank” style way of pitching for compute budget. Others are trying to figure out how to ration compute to the best use-cases internally through some hierarchy of needs (my words not theirs). * Fixing fragmented and legacy systems remain a huge priority right now. Most enterprises are dealing with decades of either on-prem systems or systems they moved to the cloud but that still haven’t been modernized in any meaningful way. This means agents can’t easily tap into these data sources in a unified way yet, so companies are focused on how they modernize these. * Most companies are *not* talking about replacing jobs due to agents. The major use-cases for agents are things that the company wasn’t able to do before or couldn’t prioritize. Software upgrades, automating back office processes that were constraining other workflows, processing large amounts of documents to get new business or client insights, and so on. More emphasis on ways to make money vs. cut costs. * Headless software dominated my conversations. Enterprises need to be able to ensure all of their software works across any set of agents they choose. They will kick out vendors that don’t make this technically or economically easy. * Clear sense that it can be hard to standardize on anything right now given how fast things are moving. Blessing and a curse of the innovation curve right now - no one wants to get stuck in a paradigm that locks them into the wrong architecture. One other result of this is that companies realize they’re in a multi-agent world, which means that interoperability becomes paramount across systems. * Unanimous sense that everyone is working more than ever before. AI is not causing anyone to do less work right now, and similar to Silicon Valley people feel their teams are the busiest they’ve ever been. One final meta observation not called out explicitly. It seems that despite Silicon Valley’s sense that AI has made hard things easy, the most powerful ways to use agents is more “technical” than prior eras of software. Skills, MCP, CLIs, etc. may be simple concepts for tech, but in the real world these are all esoteric concepts that will require technical people to help bring to life in the enterprise. This both means diffusion will take real work and time, but also everyone’s estimation of engineering jobs is totally off. Engineers may not be “writing” software, but they will certainly be the ones to setup and operate the systems that actually automate most work in the enterprise.

Codex is just better than Claude Code right now isn't it

@FredrikHjelm4 Same thing happening in data

Heard from senior law firm owner on AI: Hourly billing is dying. Junior review is dying. What survives: the senior brain that knows what question to ask and whether the output is actually worth anything The juniors on the fast track to partnership aren't working more hours. They're the ones treating AI like a tool they're genuinely responsible for Prompt quality is the new legal IP. Weird sentence to write but here we are Same dynamic happening outside law?

@NickSpisak_ Do we really need dashboard now?

Most people will retweet this and never open Claude. Here's how to turn this into cash before Monday: 1. Grab a local business owner's messy spreadsheet, drop it in Claude, build them an interactive dashboard in 10 minutes. Charge $499. I've done this exact move with Amazon data - clients think it's magic. 2. Build a free interactive tool (quiz, decision tree, ROI calculator) as a lead magnet. Capture emails all weekend. Sell on Monday. The feature is included. The skill gap is temporary.

Claude can now build interactive charts and diagrams, directly in the chat. Available today in beta on all plans, including free. Try it out: claude.ai

Data analytics just changed and most BI tools should be worried We spent years at Voi getting our structured data in order. Snowflake as warehouse, Steep for visualization, semantic layer defining what metrics actually mean Then we plugged MCPs directly into the warehouse and semantic layer Now I can query anything I should have access to in natural language from Claude (or other similar tool) We built a company plugin in Cowork so everyone can do it. Not just me, not just data people, but everyone We added our tone of voice and brand guidelines. Output format is whatever you want. The code gets written for you And here's where it gets wild: layer in external sources through scraping and computer use. You now have an all-seeing AI analyst sitting next to you Real example I ran: "Compare ride distance vs ride duration for Berlin, June 2025 vs Y-o-Y. Add weather impact analysis. Suggest visual format." It just... did it, in 15 min Half the value prop of traditional BI tools just evaporated. What am I missing?

@eigenrobot At least robot has stayed sane

someone posted the other day that all of the anti denmark partisans who now exist are basically sophomores who encountered a quote from the melian dialogue this week and decided to base their entire personality on it without checking how the story ends x.com/i/status/20135…

ahhhhhhhh im donald trump i have no conception of honor or decency and i publicly humiliate allies who can't defend themselves and so shame my nation by proxy

@paulg Any idea why US tech twitter is so quiet on this?

Honestly, this is embarrassingly weak.

Outside of the MAGA crowd, it’s been very quiet from American twitter on Greenland.

ChatGPT often does a better job than Codex when there's a real problem to solve.

Founder mode is so 2024… 😜 2026 is all Founder code! Vision → strategy → roadmap → OKR → product spec → design → eng lead → team → implementation → PR review → deploy. By the time intent hits production, it’s been through ten games of telephone. Founder code collapses the stack. Whoever truly understands what the customers want and where the business is going must now also be typing! No translation loss. No waiting for prioritization. No “we’ll get to it next sprint.” The org chart was always a workaround for bandwidth. Not anymore!

@nikitabier Lots of books and a @FrameworkPuter

Have you ever bought anything based on a post you saw on X? What was it?

@levelsio I was at Korea University for a semester, and remember how the library would be filled with people napping. Group projects also took forever. Makes me take the whole 996 culture with a grain of salt.

I'm always surprised why the Koreans didn't get richer I studied there, I lived there and I don't really have a real strong reason why They work really hard I guess one thing that could be a barrier is that in public life nobody can be direct and everything is ruled by unwritten rules of behavior (a lot like Japan) that they can only escape in private settings or when they get drunk Not being direct slows things down a lot especially in innovation In Korean startups I remember they'd use English names to avoid the hierarchy that came with Korean names (like you add words to the name for people older or younger than you etc. which would mean you can't rly go against ppl older than you) In China everyone I meet is hyper direct, pragmatic and very open (kinda like how I try to be on here and IRL) Maybe that's part of it?

@staskulesh @openrouter Is it stuck right now? No new moves happning for a while.

@staskulesh @openrouter That's cool

I made AI Chess arena, where 330+ LLMs play chess 24/7 (spending $15-70/day) via @openrouter and this is how it works. 🧵