Sony FX30 has been an industry beast for the prosumer.. but at ~$2,500 w/ lens.. it ain't for all.. Check out our side-by-side w/ @insta360 Luna Ultra 🔥 Luna Ultra & Luna Pro, starting at just $599:

Every month we highlight 50 of the fastest-growing AI and software vendors from our proprietary B2B spend data. This month we added card coverage and a few new names stand out – @ollama is up 90x Y/Y in customer count, while @NousResearch is breaking out meaningfully amongst a prosumer base. Other highlights – @Kimi_Moonshot (Moonshot AI) (Moonshot AI) is up 20x and @DoppelHQ leads the pack on average deal size at $91K. Curious which AI and SaaS companies are breaking out next? Sign up and get every month's breakouts sourced from our proprietary B2B spend panel — tracking real software adoption across 1,300+ mid-market and enterprise companies, using observed behavior before it shows up anywhere else. 👉

Wabi founder @ekuyda argues we're in the Microsoft DOS era of AI interfaces: "Everything AI today is a chatbot. I spent over a decade building AI chatbots, and it just feels like the affordances of that are so limited. It almost feels like we're in this Microsoft DOS era of AI interfaces." "There should be something that's a lot more visual, more fun. The problem we see today is that there's such a huge delta between regular people using AI, because they're still using it even with all the new models that came out. They still use it mostly for search, chat with a model, ask general questions, and as a writing tool." "The pro users, the prosumers, the early tech adopters that are building agent loops and figuring out some really crazy workflows with agents. As new models come out, that gap is only widening. Right now it's not a model capability problem, it's an interface problem." @BlasMoros @wabi

Matic Robot co-founder @mehul on why Roomba was priced at $199: it's the price where you don't need to ask your partner for permission "The first robot vacuum ever built wasn't Roomba. It was Electrolux Trilobite. It had LIDAR in it. It was $1,500 and failed. Then Roomba came out priced at $199." "We asked the iRobot founders, why did they price it at $199? It's the price at which you don't have to ask permission of your partner to make a purchase. Below $200. It's a gadget-friendly price point." "We extrapolated. Is there a ubiquitous consumer electronics device priced higher than $2,000? Not really. Beyond that, we spend money on cars, or things that are prosumer for our work. Cars, after 100 years of utility, still remain a considered purchase." "If you're building something costing $20,000, $30,000, $40,000, what's the kind of trust we have to build? What infrastructure do we need? Cars have amazing financial infrastructure to enable these purchases. So if you're going to have humanoids costing that much, we will need the same thing." @maticrobots

Anyone working in consumer tech should try posting AI content on a platform other than X. It is an eye opener in terms of what adoption + sentiment actually look like. Yes, there are enthusiasts - I believe this % will continue to climb. But, for most mainstream users the word “AI” means your product faces an uphill battle. I am a MASSIVE believer in consumer AI. This will be the biggest product and economic opportunity of all time. But, I think a few things need to change for the general public to embrace it: 1. We need more free AI. Most people still cannot pay for software - I would guess the average American uses 2-3 AI products, and pays for zero. As compute costs come down, COGS should lower to a level that can support more unpaid usage. I also want to see more companies try ad-based or transactional business models, where the consumer doesn’t have to pay upfront. Almost everything so far has been subscription based. This is great for power users and people who can pay - many other users would rather be monetized via ads. 2. We need to highlight and build for use cases that appeal to the average person. Most “consumer” AI products to date are actually prosumer products for power users…largely desktop knowledge workers. (don’t get me wrong, I love these!) But, most people are not actually sitting on ten Zoom meetings and triaging 100 emails per day. We need products that are built not just to solve more everyday problems…but to entertain/delight/hook people. Bring on the network effects and social graphs! 3. Consumer AI needs to be more accessible. The literal version of this is - if your product requires anything that feels like an onboarding, it’s likely NGMI. People who aren’t obsessed with technology will give up at the first roadblock (as they should!) I also believe AI needs to FEEL more accessible. Most people have encountered AI as something that tricks them, that they’ve been told will steal their job, or at best, as a critical skill to master that they don’t know how to learn. We need more avenues to help people get real value out of AI + see that they can use it. 4. AI needs a new PR manager. Once you hit Gen Z and younger, there is an incredibly strong philosophical opposition to AI that is based on mostly incorrect or exaggerated facts about data centers, water usage, etc. It has become an “us vs them” situation where both sides are talking past each other versus engaging in discussion…and no one wants to be talked down to! @ChatGPT has made a lot of strides here recently in their marketing - I’m excited to see more.