January 20, 2025 4 min to read

2025 wk4: AI Wars & Wardley Maps 🗺️

Tech, AI, and a dash of politics. Plus, Wardley Maps for the win! 🗺️

Note: These are my personal views, not those of my employer.

Cool Tech

• This morning my buddy 🔗👔 Daniel told me about Rahoot, an OSS clone of Kahoot he runs on Cloud Run. He also told me he consumes Hacker News from here.

• Google stock (GOOG) exceeded maxint this week (above 200$). Alphabet price target raised to $230 from $217 at BMO Capital. You might not care. I do.

Cool people

• My former colleague Jos Visser (“Zxoogler”), now at OpenAI, came to Google Zurich for a talk. His only CTA was a nice Wednesday wisdom blog. Strongly opinionated, as we like it.

• My former colleague Brian Grant doesn’t need intros: he overlooked k8s expansion and execution, CRDs, surface, IaC and so on. He’s basically a genius. I follow him on Medium through ITNEXT and I think he is one of the few to have ideas which can take Terrafom/Pulumi/friends to the next level. Read more from him here.

Politics

• 22jan. Elon Musk and Sam Altman take to social media to fight over Stargate. “Musk is spearheading the Department of Government Efficiency (DOGE), a U.S. government advisory commission recommending deep cuts to federal agencies. DOGE, first announced last year, was made more official Monday by President Donald Trump’s executive order — but the commission faces a series of legal challenges.” Now, the Italian equivalent DOGE (“Ministero della semplificazione normativa”) was introduced - if memory serves me - by Lega Nord and Calderoli under Berlusconi. Which is to say, The fruit doesn’t fall very far away from the tree.

Gen AI Wars

Note: I’m a Googler, I might be biased. Some of these links come from my email subscriptions to substack.

• Google agrees to new $1 billion investment in Anthropic - Google invests further $1bn in OpenAI rival Anthropic
• Microsoft News Roundup: Nightmare week of price hikes, rebrands, and remembering how Bill Gates lost $400 billion. Whoopsie.
• Google releases free Gemini 2.0 Flash Thinking model, pressuring OpenAI’s premium strategy.
• Google’s Gemini is already winning the next-gen assistant wars. Looking forward to it. My apartment is full of Google Wifi, google Nest, and my kids love the nice compact pink Google Chromecast remote (we call it pinkie).
• OpenAI starts the release of its o3 mini AI model for ChatGPT, and it’s got a nice speed boost over o1. Nice!
• Anthropic chief says AI could surpass “almost all humans at almost everything” shortly after 2027. Strangely enough, and probably nobody cares, but I concur. Speaking at Journal House in Davos, Amodei said, “I don’t know exactly when it’ll come, I don’t know if it’ll be 2027. I think it’s plausible it could be longer than that. I don’t think it will be a whole bunch longer than that when AI systems are better than humans at almost everything. Better than almost all humans at almost everything. And then eventually better than all humans at everything, even robotics.” Dario, Davos to zurich is a 2h train. Come over, I’ll buy you a coffee.
• TikTok owner ByteDance plans to spend $12 bln on AI chips in 2025 (Reuters). Let’s see what happens in the AI chip war between China, US and.. Holland. ;)
• Perplexity launches Sonar, an API for AI search (TechCrunch).
• Anthropic plans to release a ‘two-way’ voice mode for Claude (TechCrunch)
• OpenAI tells India court ChatGPT data removal will breach US legal obligations. Smart 🍑

🇬🇧/🇺🇸 English idiosyncrasies

When you work for an American company, you’re proud of being a linguist (and a grammar Nazi - no, not like this one), you find a few words a bit awkward.

• move the meeting 1h back is quite counterintuitively moving in the future. Why? I love the explanation here.
• Biweekly. As with “biweekly” (can be twice a week or every other week depending on the speaker/listener) there is definitely some ambiguity.
• Idiots don’t sync. Just messing with you (and proving this content was not AI - generated). This is NOT the meaning of idiosyncrasy.

‍💼 Strategy

• Ever heard of Simon Wardley Maps? If not check his creator’s Blah talk. I was just talking to my friend Steve McGhee with whom I was discussing about Wardley map for SRE/Reliability a few years ago. Create your fake strategy here (on Heroku)! I’ve started working on a Wardley Map or Ops and Agents in the Gen AI / LLM space.

Gen AI

• Vulnhuntr allows you tu use Gen AI to find vulnerabilities in your code. Thanks Davide Guerri for the tip.