--- type: "article" title: "Weekly Links: AI Marketing Disasters, Anthropic IPO, and Tips for Agentic Engineering" summary: "Anthropic files for IPO and also suggests the world should slow down AI research. High profile errors at Starbucks and Meta suggest the opposite." newsletter: "Steampunk AI" newsletter_handle: "steampunkai" newsletter_url: "https://usecommune.com/n/steampunkai" author: "Steven Willmott (@njyx)" published: "2026-06-07T08:14:58.000Z" canonical_url: "https://usecommune.com/n/steampunkai/a/weekly-links-ai-marketing-disasters-anthropic-ipo-and-tips-f" markdown_url: "https://usecommune.com/n/steampunkai/a/weekly-links-ai-marketing-disasters-anthropic-ipo-and-tips-f.md" chat_url: "https://usecommune.com/n/steampunkai/a/weekly-links-ai-marketing-disasters-anthropic-ipo-and-tips-f/chat" source_url: "https://www.steampunkai.com/weekly-links-ai-marketing-disasters-anthropic-ipo-and-tips-for-agentic-engineering/" body_source: "imported" likes: 3 replies: 5 body_words: 486 --- # Weekly Links: AI Marketing Disasters, Anthropic IPO, and Tips for Agentic Engineering This week, Agents [get therapists](https://www.agenttherapy.live/), [Supabase hits 10.5B in valuation](https://www.cnbc.com/2026/06/04/database-startup-supabase-raises-500-million-10point5-billion-valuation.html), and Amazon product search [begins to invent products you can't buy](https://www.theverge.com/tech/942547/amazon-search-bar-ai-images) (yes, it's deliberate). Finally, Anthropic suggests the [world pause AI innovation](https://www.theguardian.com/technology/2026/jun/05/anthropic-urges-temporary-pause-on-ai-development-to-discuss-risks), but in the same week, [raises $65B](https://www.anthropic.com/news/series-h) and [files for an IPO](https://www.wsj.com/tech/anthropic-ipo-paperwork-9a48c35e). On to the main stories this week: - [How a Starbucks marketing stunt spiraled into mass boycotts in South Korea](https://www.theguardian.com/world/2026/jun/06/starbucks-south-korea-tank-day-promotion-blunder). This is a wild story, and the key to it all: "Marketers chose the slogan after consulting an AI tool, looking for suggestions, Shinsegae Group said. It turned out some managers who approved the campaign never opened the email attachments showing the marketing material." Yes, the campaign was "relevant" to the date, but not in the way you would have hoped. - [The Newest Instagram "Exploit" is the Goofiest I've Seen](https://www.0xsid.com/blog/meta-account-takeover-fiasco). Not to be left out, Meta had a serious incident this week with a new AI system for handling user account resets. The security situation is laughable, and the exploit [still is not properly fixed](https://bregg.com/post.php?slug=meta-instagram-ai-fix-incomplete-vulnerability-continues). In fact, the fix itself smacks of AI coding (the button for triggering account help has been removed, but API access for the same thing remains). It's scary that even the biggest companies seemingly don't know how to triage functions that are critical and security-related from run-of-the-mill functionality. - [How AI has de-skilled translation](https://www.ft.com/content/75bc43ed-6a57-4f77-929c-83238f5cfe60?syn-25a6b1a6=1&shem=dsdf,sharefoc,agadiscoversdl,,sh/x/discover/m1/4). This week's column rapidly risks turning into a full-on "AI is terrible" edition, but I wanted to include this because it's a clear illustration of how some job roles are being eliminated. In translation, in particular, first Google Translate and now LLMs have gotten so good that human adjustment is barely needed for some types of work. Increasingly, translation jobs are becoming "fix what the AI missed," and understandably, the most skilled translators are choosing to step out of the market altogether rather than do this. - [NVIDIA Launches Cosmos 3, the Open Frontier Foundation Model for Physical AI](https://nvidianews.nvidia.com/news/nvidia-launches-cosmos-3-the-open-frontier-foundation-model-for-physical-ai). Accelerating the trend for all-in-one models that bake in an understanding of the world, NVIDIA's new open-source model combines a physical world model with an LLM and a framework for action-taking. This is the current path many robotics companies are betting on to enable physical AI. It's interesting to see NVIDIA so engaged in this, but logical since physical AI may end up being even bigger than data center AI. - [BDD, ADR, PRD, WTF: Capturing Decisions for Humans and AI Alike — Michal Cichra](https://www.youtube.com/watch?v=504PvfXou5Y). My Safe Intelligence colleague has a great talk up on the AI.Engineer YouTube that is getting a lot of love. The key point: if you're going to use Agents to code, you need to rethink how you communicate boundaries to them. Definitely worth checking out! [https://www.youtube.com/@aiDotEngineer](https://www.youtube.com/@aiDotEngineer)Lastly, this week: if you're a LEGO fan, you might love [this](https://www.dezeen.com/2026/06/04/lego-sagrada-familia-largest-set-antoni-gaudi/?shem=dsdf,sharefoc,agadiscoversdl,,sh/x/discover/m1/4) - a 12,060-piece set of the Sagrada Família in Barcelona. Thank you to various colleagues for surfacing great links this week! *** ## Discussion (5 replies) Numbering shows the reply tree: `2.1` is a reply to `2`. Message text is reproduced verbatim as posted. ### 1. Fran Mendez (@fmvilas) - 2026-06-07T08:52:56.861Z Dude too many laughable stories this week 🤣 Actually, as we say in Spain, "me rio por no llorar". It ultimately comes down humans trusting AI so much that they don't review the output anymore. It's tempting. I'm not going to say I never did it. But mate, a whole marketing campaign and a fix that consists of removing the UI only... WHO'S OPERATING THAT AI! 😅 ### 2. Steven Willmott (@njyx) - 2026-06-08T06:58:49.618Z Ha ha -- yes, this week really had some crazy "yolo AI" ones. Somehow the button to implement an AI output into the real world needs to be "95% this is pretty good, 5% if will destroy your business - Y/N" ### 3. Joyce Stack (@joycestack) - 2026-06-08T08:30:35.161Z Perfect timing with the ADR, BDD, TDD link thank you. I spent last week building a RAG over ADRs in one of our org repos. They told us very little interesting information TBH. I'm curious about what role BDD, TDD and ADRs have in this new world given that a spec is now a also a decision record. I have also seen how some specs are generating lots of ADRs about every brain fart it has. ### 4. Steven Willmott (@njyx) - 2026-06-08T10:16:23.511Z Ha ha - nice coincidence. I think with ADRs indeed it can out of control too. We actually don't have many of those though (try to use sparingly) and iterate on them a lot. So being really picky on what they say. I think a big problem is going to be versioning all these things in sync. So if you decide to change the product how to do you go down the whole stack and amend everything. To some extent this has always been a problem but velocity was not high enough to burn you too much. ### 5. Joyce Stack (@joycestack) - 2026-06-08T18:02:40.387Z > Quoting Steven Willmott (@njyx): Ha ha - nice coincidence. I think with ADRs indeed it can out of control too. We actually don't have many of those though (try to use sparingly) and iterate on them a lot. So being really picky on wha… That is interesting re: ADRs. I pulled out a dashboard of "what questions can be answered" and due to the fact that there is a lack of discipline, linting, templating etc then there was no way of getting a sensible answer to something like "why are teams using X DB over Y DB?". No supersedes means that you have no idea of what changed for teams to move to Y DB rather than X. Another pointless process over outcomes finding.