Inklings #001 📧
Yes, yes, another new title and format tweak. But this time, I'm fairly confident in the path forward here. As I've said from the get-go, the aim of the newsletter component of Spyglass is for it to be a less formal, more parseable delivery mechanism of links and thoughts. And as the Spyglass columns have crept longer, so too has the newsletter, as the email size-limit overflow indicators have made me acutely aware of over time. So I'm trying to whittle it back a bit to the basics. I aim for it to be something simple to open, scan, digest, and close. And then repeat on regular intervals at regular times.
Thoughts On...
💫 Cursor + SpaceX – In noting a compute deal between the two last week, I wrote, "Would be very curious how this deal came together/was structured" – because while the high level notion made some level of sense, it sure felt like there would need to be a lot of structure around it for both sides. As it turns out, that structure is a $60B call option for SpaceX to buy the entire company! And if they don't, they'll pay a "mere" $10B – sort of a de-facto break-up fee, albeit tied to compute deals. Price aside, this actually makes more sense to me. Cursor is under immense pressure from the foundation labs who have decided their space is the most important one to own at the moment. They clearly had an option on the table to keep going, which seems to say a lot (though also perhaps bullishness around SpaceX pre-IPO shares, obviously). And SpaceX – meaning, the xAI subsidiary – is under immense pressure because they can't compete in that um, space, yet. And why buy two utters when you can get the whole cow? Next question: will Anthropic and/or OpenAI now fully pull their models from Cursor? Sure, that means forgoing money, but Anthropic in particular could undoubtedly use the capacity back at the moment. The real loser here may be Meta, which has no viable coding option yet. Forget Space Twitter, and Space Data Centers, now we have Space Vibe Coding! [NYT]
🐦 Xitter Custom Timelines – This feels a lot like a feature created to try to put Grok to use rather than any sort of actual user demand. Xitter would undoubtedly say this is the holy grail for casual users, but I think it's sort of the opposite: they want one feed built for them, not 75 they have to hunt and peck between. But really, are "casual" users still even on Xitter? I guess they want to try to lure them, and the growth hacks seem to be doing that to some extent, but this would also seem to be an indicator that they still don't stick around. The service still has an edge in breaking news (and sports) but Threads now eats their lunch algorithmically for those non-vital times. If Meta can bridge that gap... Xitter is increasingly a mess content-wise. They keep insisting links aren't being "de-boosted" (anymore), but then they do things like jack up the price of posting links by those using the APIs (so, news orgs) by – checks notes – 1,900%. But that's just to combat spam, you see. How about you task Grok with that. Mainly, it's nice to see that from Twitter to X, the people building the product are continuing the legacy of not understanding what the user base actually wants and instead focusing on random stuff to try to win battles they lost years ago. [@nikitabier]
🖼️ ChatGPT Images 2.0 – It's good. Very good. Does it eat 'Nano Banana'? Time will tell – presumably, Google has an answer in the works... will they save it for Google I/O in a month or move it up? – but in general, it's good to see OpenAI back on offense, product-wise after the past few months playing full-on defense. And it's obviously smart to play up image capabilities as it's something that Anthropic simply does not offer with Claude. You might say that focus has helped them – and certainly, OpenAI was doing far too much a few months ago – but image generation feels like a fairly core competency – including in the enterprise, by the way – that Anthropic will likely have to match at some point. For now, it's OpenAI vs. Google (and yes, xAI, Microsoft, Midjourney, Black Forest Labs, and others to lesser extents), and it feels like OpenAI is back out front again. (Though I do miss the 'DALL-E' name – we've come a long way – it seems like this is a new set of models, including with "Thinking" and web search capabilities.) [TechCrunch]
🍏 Tim Cook's "Textbook Succession" – One focal point of my piece about Cook's transition from CEO was "why now?" and in his town hall with Apple employees, he seemingly confirmed the notion that it was a convergence of factors, notably: the business doing great, a robust pipeline of products, and John Ternus being ready. I'm still guessing the 50th anniversary of the company was at play too (as well as his 15th anniversary as CEO, which he'll pass in late August just before he steps down – and the fact that he'll get to perform his new role on the Board for a nice, round 10 years, if Apple still cares about the 75 age limit at that point), but I do generally believe the blockbuster Q1 earnings (and the market cap being back above $4T – a proxy for investor confidence) with so much uncertainty looming made this a sort of a perfect storm, in a good way for Cook. You don't want to exit when everything is going to hell – just ask Bob Iger, which (as a former Apple board member) Cook probably did! In his remarks, Cook also clearly wanted to get in front of any health whispers – which as I noted in a footnote, have been out there, but seemingly not addressed... I have many more thoughts on the Ternus side of this equation. Likely for a post tomorrow. [Bloomberg 🔒]
Below, members of The Inner Ring will find thoughts on:
• OpenAI's Forward-Deployment Company
• Amazon's $25B Anthropic Investment
• Google Falling Behind in AI Coding
• and more...