Figure breaks up with OpenAI

Plus: Sonos streaming; EA rollercoaster

Author

Joey Padgett
February 06, 2025

In partnership with

Welcome back to Forests Over Trees, your weekly tech strategy newsletter. It’s time to zoom-out, connect dots, and (try to) predict the future.

Here’s the plan:

  • Tech News Takes — super-short analysis and commentary

  • Tool of the Week — tools you’ll find useful

  • Strategy Tips — strategy nuggets (for business and life)

  • F/T Shoutouts — sharing launches, tech events, and other reads

Figure breaks up with OpenAI

Plus: Sonos streaming; EA rollercoaster

⚡ Tech News Takes ⚡

  • What’s up: In a few months, Sonos is launching their first streaming device, codenamed Pinewood, expected to cost $200–$400. Like competitor devices, you’ll be able to access multiple streaming apps (Netflix, Max, Disney+, etc.). And Sonos will let customers connect it to their other Sonos speakers for surround sound. They are trying to rebound after a bleak 2024, which saw their former CEO resign after a particularly poorly-received app redesign.

  • So what: A few thoughts here. First, this seems like a good way to diversify revenue. Today, hardware is 95% of their business. If they can imitate Roku, Amazon, and Apple, selling ads in the form of paid placements and/or taking a cut of subscription sign-ups, that could go a long way. Second, it seems it seems like a smart move to focus on building new, valuable things that some users will pay for vs. the thankless, time-consuming app redesign they prioritized in 2024.

  • What’s up: Figure AI, the robotics startup working on general-purpose humanoid robots, is ending its OpenAI partnership to focus on its own AI models. Figure hinted at a “major breakthrough” for its in-house AI and will unveil new capabilities within 30 days. Figure has raised $1.5B from investors including OpenAI, and its last raise was at a $2.6B valuation. According to context provided by CEO Brett Adcock to TechCrunch, “We found that to solve embodied AI at scale in the real world, you have to vertically integrate robot AI…we can’t outsource AI for the same reason we can’t outsource our hardware.”

  • So what: I agree with Adcock that there’s benefits to vertically integrating hardware and software (like Apple has done), but when you publicly break-up with OpenAI, one of the hottest companies around… there’s something else going on. Two guesses: 1/ OpenAI is too focused on its consumer use cases and not dedicating enough time/energy (from Figure’s POV) to robot AI. 2/ OpenAI is not sufficiently exclusive, and Figure might lose out to competitors supported by OpenAI or to a homegrown OpenAI robot itself. There seems to be more evidence for #2 (OpenAI is working with Figure competitor “1X” and also filed their own patent for a humanoid robot).

  • What’s up: EA recently lowered their revenue forecasts, pointing to weak performance by 1/ their latest installment of the soccer franchise FC 25 and 2/ a new game called Dragon Age. But they also announced a $1B share repurchase, and the CEO said he’s “confident in a return to growth in FY26.” His optimism is partially due to their planned release of the next installment of Battlefield, another popular franchise.

  • So what: From an outsider’s perspective, the share buyback seems completely oriented towards appeasing shareholders (and taming the sting of lower forecasts), but it’s not going to actually do anything to materially help the business… And if previously popular titles like FC are on the decline, what makes them think another previously popular title (Battlefield) will be different? But in talking to a few gaming industry vets, they shared that amid this really tough period in the industry (layoffs, AAA game flops, etc.) milking old IP is one of the only things that’s actually working. So in other words, this CEO (and other gaming leaders) seem to have little choice but to keep trying what used to work and hoping for the best.

🛠️ Tool of the Week 🛠️

The #1 AI Meeting Assistant

  • Summarize 1-hour meetings into one-page recaps

  • Say goodbye to post-meeting data entry

  • Claim 30 days of unlimited AI notes today

🧭 Strategy Tips 🧭

Figure Breaks-up With OpenAI

Today's strategy tip is all about knowing whether to go it alone.

Specifically, we’ll analyze Figure’s decision to build its own AI models, using the Make vs. Buy framework to break it down. But first, let’s meet the framework.

Make vs. Buy 101

Companies constantly have to decide whether to build things in-house (aka “make”) or outsource them to someone else (aka “buy”).

And for small things that aren’t core to the business, it should be an automatic buy… If you’re opening a restaurant – you should be buying (not making!) toilet paper.

But for things that do matter to the business, make vs. buy comes down to cost, differentiation, and control.

  • Make: higher cost, better differentiation, more independence/control

  • Buy: lower cost, more cookie-cutter, increased vendor dependence

With that context, let’s apply the framework to Figure.

Why Figure Chose “Make”

  1. AI is core for Figure – First, it’s clear that AI for an autonomous robot is mission-critical, so it’s not an automatic buy. It’s perfectly logical to make an active decision.

  2. Control boosts performance – Being able to tailor their AI/software to the specifics of their robots/hardware will increase performance.

  3. Dependence on OpenAI is risky – OpenAI partnered with Figure competitors, and might be distracted by the litany of other AI projects/priorities they’re working on.  

Figure’s CEO hinted that a major announcement was imminent, so I’m curious to see how real/advanced their homegrown AI is. Only then will we get an idea of whether they’ve made the right make vs. buy call…

🌲 F/T Shoutouts 🌲

  • A Treasure Trove of Gaming Analysis – while those gaming vets were schooling me on EA’s situation, they also pointed me to this incredible, insight-filled deck from industry analyst Matthew Ball. Here’s a snip of a mind-boggling graph from the deck:

  • Founder Lessons — this podcast is nominally an interview of a legendary markets investor, but secretly it’s just a ton of lessons for founders. If you’re building something, I think you’ll dig it.

Want to suggest a shoutout? Send it here.

The forest is growing.
Feel free to share this post.

To connect or collaborate, please reply to this email or send me a note.