A video about the new paintings I have made for Minecraft youtu.be/x5roenGT1QI?si=A_df…

Out now: “Trusted journalism in the age of generative AI.” A report researched and written by @_FelixSimon_ , @OlleZ , Kati Bremme and myself, published by the @EBU_HQ #AI #media

Sorry to break it to you. It's a weird guy in a fez holding a stick. One of the strangest paintings I've made. I remember painting it when I went to the Royal College of Art. Maybe 1997??

💬 One of our panel sessions this afternoon discussed values & ethics in the era of #AI ❓ How do you feel about the AI-powered future in #PublicServiceMedia? (71% of our audience @CERN feel positive!) 👀 @Dr_JMRickli @kurczabinska @monikagarba @OlleZ @kati_bremme @AlexBritoRTP

My 🌶️ spicy post-thoughts after last week's AI panel discussion at @SvTidskrifter together with @chrismoranuk (@guardian), @AgnesStenbom (@SchibstedGroup), and @OlleZ (@sverigesradio). What was I doing on the same stage as the head of AI and innovations at these top-tier media houses? I see my role as the counterbalance to what easily becomes group thinking within an industry. Since I’m an engineer with a business mindset and no dog in the game, my comments are purely made from a technical and business perspective. Since we only had 25 minutes on stage, here are my post-thoughts we didn’t have time to discuss. ## TRANSPARENCY My brain translates “transparency” to “journalists want to make sure that content written by an AI, is clearly labeled as AI-generated”. The same discussion was probably taking place in the late 18th century when, instead of items being produced by hand, items started to be produced by machines. Here’s a challenge: Look around and point at handmade things. There is not much to point at. Machines make most items. This is not controversial (anymore). ## RISK AND HUMAN-IN-THE-LOOP The example on stage being a newspaper in Norway incorrectly reported that a famous football player was shot dead. The article was generated by an AI and published without a “human-in-the-loop.” This was obviously a mistake, but mistakes alone are not a reason to stop progress. The same phenomenon happens in the autonomous car industry. Tesla now has proof that their autonomous cars are statistically safer than human-controlled cars. But, whenever a Tesla car on autopilot is involved in a car crash, people push to halt its use. Progress must be approached with caution — but not stifled by fear. ## WHAT HAVE YOU ACCOMPLISHED? I’ve heard people answer this question a lot lately — both in a journalistic context and elsewhere. To my surprise, 75% of the answers are a variation of “We’ve created a policy describing what we’re not allowed to do with AI.” EU regulators would be proud... ## CONCLUSION - Don't fear machines. Use them as a tool to 10x yourself. - Don't minimize risk. Some rockets will blow up before we reach Mars. - Ignore people who write "policy documents." Focus on actually doing.