John Bonten

Tech Operator. Co-Founder & Investor SpotiAngels. Formerly International @Spotify

[John Bonten @johnbonten]

Cat's out of the bag! Thanks @Siftedeu for reporting on our angel investing initiative. Excited to bring back the band together with @katiayakov and @bringe! Say hi and get in touch via https://t.co/7LoGsrt1rz to learn more.

[Sifted @Siftedeu]

A trio of Spotify alums have launched a new angel syndicate in Berlin.

Now, Spoti Angels will invest between €25k to €150k in European pre-seed and seed stage startups in consumer tech, the creator economy and the future of work. 💰
https://t.co/haS761R5RD

[John Bonten @johnbonten]

Huge congrats!

[Passionfroot @passionfrootme]

We’re excited to announce our $15M Series A, led by @insightpartners.

Passionfroot is the execution layer for creator-led growth — the way the next generation of companies reach the market.

As AI makes software easy to build, distribution becomes the real moat. Buyers no longer discover products through ads or SEO; they find them through the independent voices they already trust — the creators, operators and builders they follow. Passionfroot makes working with them something any team can actually run, end to end: discovery, contracts, campaigns and global payments, unified by AI.

Today we power creator-led growth for Figma, Replit, Airtable, Luma and countless category-leading AI-natives and Fortune 500s.

This raise lets us open our New York office, invest deeply in the product, and grow the team across New York, Berlin and São Paulo.

Huge thanks to Insight Partners for leading the round, to Creandum, Supernode and s16vc for doubling down, and to our customers for building this with us.

Read the story in TechCrunch 👇

[Photoroom @photoroom_ML]

🚀 Meet PRX Pixel.
Our new open-source 7B text-to-image model that generates images directly in pixel space.
After months of pretraining on hundreds of millions of images, supervised fine-tuning, and preference alignment, we're excited to share a first public preview.
The weights are already available, and we're currently working on integrating the model directly into Diffusers 🤗to make the model even easier to use.
Test it yourself in the demo below. And as always, we'll be sharing the full story behind the model through a series of technical blog posts covering the entire training recipe.
Link in the comments 👇

[John Bonten @johnbonten]

@aurexcoder0 nano

[John Bonten @johnbonten]

@Oliviacoder1 Agent

[Matt Rouif @matthieurouif]

🚀 We’re launching something big today! With #GPTImage from @OpenAI, Photoroom just rolled out three new tools to help sellers and small teams create pro-quality visuals in seconds—no studio required.

Photoroom mission is to make the power of imagery accessible to all entrepreneurs, all commerce. We are taking this mission one step further

[Aaron Levie @levie]

Identifying and building out sustainable moats in AI is one of the most important and exciting topics if you're in software right now.

Given the fact that there are multiple at-scale frontier AI model providers with effectively infinite budgets, training data appears to be broadly available, and research breakthroughs spread relatively quickly through the community, we can anticipate that the cutthroat competition continues to drive down underlying AI pricing and increases performance over time. This is great news for developers.

One implication, however, is that we can expect the AI models themselves to handle more and more of the logic that customers are looking for directly; and it means that each individual market will see increased competition as the barriers to entry go down over time.

This means that the best problems to go after are ones where there is a major distance between the customer and the underlying AI model. And filling that gap must be sustainable, differentiated, and something people need to pay for. This value can show up in a variety of ways, but it will most likely correlate at least to some degree to the amount of software you've created and the uniqueness of the data set you touch or manage.

As with any prior software markets, this means that it's better to go after problems that can both start deceptively simply --so it's quick and easy to get going and see if you have product market fit-- but that have an almost endlessly complex surface area to cover to really solve them.

These opportunities will often be found in verticals that few other players understand, in integrating across many systems that are hard to connect to, where there are deep workflows that are challenging to get right (no matter what breakthroughs occur at the AI layer), and where the value increases the more data is in the system. Ultimately, the more frustrating and annoying the problem to build for, the more sustainable the moat.

We're still in the early innings of understanding how AI moats will get built over time, but it's easily the most dynamic and interesting period we've ever had in software.

[Khaled Helioui @KHelioui]

640m MAUs, 252m paying subscribers, double digit revenue growth to get to c. €4bn quarterly Revenue while expanding margins. @Spotify is about to become the first European startup of this generation to make it as a $100bn+ publicly traded company. A green beacon of light amidst tired European doomerism

[Sarah Tavel @sarahtavel]

It seems inevitable that as the underlying foundation models become more powerful, the LLM players will seek to justify the enormous investment that has gone into training their models by moving "up the stack", and evolve from an API or chat interface, to async agents.

[Daniel Ek @eldsjal]

It's been 6 years since we became a public company. It’s pretty wild to think that by most metrics, Spotify is now 3x or more on where we were that day. I am reminded of Bill Gates’ quote, “Most people overestimate what they can do in one year and underestimate what they can do in ten years.” Oh, and have a look at the amazing flag choice in the last picture. 😆🇨🇭🇸🇪

[Nico Wittenborn @ncsh]

PhotoRoom is now a Top 10 US Photo App!

Youngest company, lowest $ funding, 3rd highest rating and only EU HQ of the 10:

[Matt Rouif @matthieurouif]

We are cutting the price of Photoroom API by 90% to $0.02/image.

Companies like @wbpictures , @netflix , @AudiOfficial, @woltapp @RepublicLabelRP build with Photoroom tech and we want to make this even more accessible!

In the early days, many people initially thought Photoroom was "only" a background removal app. What they didn't understand is that the segmentation task - computer vision name for background removal - is about understanding images. Segmentation is, in fact, a building block of the internet & we are the best at it. Come build with our tech ❤️

[John Bonten @johnbonten]

💥 @photoroom_app @matthieurouif @EliotAndres

[Matt Rouif @matthieurouif]

We are THRILLED to announce a major milestone today: our Series B, raising $43M with Balderton and Aglae

We also announce 6 new GenAI features in Photoroom. More importantly, these features are powered by Photoroom's foundation model, the best model for commerce photography.📸✨

[John Bonten @johnbonten]

@blixt @framer 💚💚

[Historic Vids @historyinmemes]

Abandoned Hotel Belvédère was an iconic Swiss hotel on the edge of the Rhone Glacier. It was built in 1882 by young hotelier Josef Seiler in one of the hairpin bends of the Furka Pass, one of the snowiest regions in Switzerland. The panoramic location attracted a pampered clientele who sometimes stayed for weeks during the summers.

The hotel industry in Switzerland boomed at the turn of the 20th century, and Hotel Belvédère's popularity increased with the opening of new railway lines. However, from the 1960s, the number of guests at the hotel and other Alpine hotels decreased due to faster and more powerful cars. The once two- or three-day trip through the pass became a one-day round trip, and guests no longer stayed overnight.

The hotel closed a couple of years ago and will likely never reopen.

[John Bonten @johnbonten]

Congrats @an21m, @michaeljames and the rest of the @amoamoamo team! Beautiful execution!

[amo @amoamoamo]

You are more than a profile pic and a grid of edited photos. You are everything you wear, say, do, and love — a beautiful and chaotic mess that means you’re human.

ID by amo is a new kind of social profile that you make with your friends, on a limitless board.

No grid to curate, followers to collect, or meaningless story replies. Just you and the crew, sharing what makes you, you.

Welcome to amo. Just friends.

[Harry Stebbings @HarryStebbings]

Is @photoroom_app the most capital efficient company to come out of @ycombinator?

💵 $50M in ARR in 3 years

🔥 $20M in ARR on just $2M invested.

🌎 100M consumer downloads

That is an insane $961,538 per employee.

Top 6 lessons from our 20VC episode w/ @matthieurouif:

[Sarah Tavel @sarahtavel]

For the past 25 years, application software startups have had a singular focus: increasing company and employee productivity. AI enables a paradigm shift: Rather than sell software to improve an end-user's productivity, consider what it would look like to sell *the work itself*.

When you sell the work, instead of selling a 15% productivity improvement, you sell a 95% improvement. Incumbents are not only advantaged in selling software, they are stuck there. You don't need to play their game. Post: https://t.co/29BkdGIMC3

[John Bonten @johnbonten]

@mathieuvdpoel 🌈 🥇

[UCI @UCI_cycling]

Photo of the day 😍

#GlasgowScotland2023

[Daniel Ek @eldsjal]

One of my team members reminded me that in June of 2016, Spotify reached 100 million users. Now we have more than twice as many paying users (>200m) and around 6.5% of world population uses Spotify 🌎🤯.
… and growth last few quarters have picked up, exciting times!

[The Spectator Index @spectatorindex]

Kilauea volcano currently erupting on Hawaii's Big Island.