New solutions based on artificial intelligence dominate the business models in the youngest year of the start-up forge Y Combinator.
(Photo: REUTERS)
san francisco The start-up development assistant Y Combinator is one of the most important indicators of future trends in Silicon Valley. One topic dominates the current year: “Artificial intelligence has never been as important as it is today,” said the head of Y Combinator, Garry Tan, at the presentation of the new business models recently in San Francisco. From 20,000 applications, 282 were accepted into the program.
The text robot ChatGPT from the company OpenAI made artificial intelligence accessible to a mass audience. The software can write poems, develop business models or do homework for university studies. Other systems can produce photo quality images, produce videos or music. The collective term for these technologies is generative artificial intelligence.
Technologies have improved massively over the past few months. 61 start-ups in the Y Combinator run list generative artificial intelligence as a core of their business model. They see particularly promising applications for everyday business in these five areas.
1. Booth AI and Dream 3D: Virtual photo shoots
If you want to sell your products well, you need good and professional photos. However, choosing the right location, the right photographer and the right presentation is time-consuming and expensive. The start-up Booth AI wants to have this process taken over by artificial intelligence. All customers have to do is upload images of their product. The system takes care of the rest. For example, a photo of a jacket can be turned into an image of an AI-generated model wearing the jacket.
The start-up Dream 3D wants to go one step further. It wants to digitally display products and entire worlds in 3D. In contrast to individual photos, this should also make it possible to create entire video sequences in the future.
2. Wild Moose: Tech support with virtual helper
Technical problems can block important processes in companies and become very expensive. Finding the errors often requires difficult processes. Log entries must be evaluated or different sources of error must be excluded individually. The start-up Wild Moose wants to automate these processes.
An AI system should support professionals in the IT departments of companies so that they can fix errors as quickly as possible. While systems like Github-Copilot would help with writing software and identifying bugs, they are limited to computer code. Wild Moose takes all data and all resulting problems into focus, says company boss Yasmin Dunsky.
3. Play.ht and EzDubs: One Voice for One Brand
Computers have been able to read texts for years. But their voices mostly sound wooden and unnatural. The two start-ups Play.ht and EzDubs want to change that. Play.ht has built a database of hundreds of voices in more than 130 languages. These should not only help texts read aloud by computers to sound better, they should also be able to give companies their own voice.
The company calls this process voice cloning. The head of the company or the head of customer service can, for example, have their voice recorded digitally so that they can reproduce the company’s messages in their own words. In the future, companies might not only need a logo, but also their own voice.
More about artificial intelligence
EzDubs, on the other hand, focuses on a specific problem: translating videos into multiple languages. The company’s software specializes in live translation of videos. Thanks to this technology, a company can broadcast an event live via YouTube, and the voices are played out in real time in several languages.
Among other things, the company uses its software so that interviews with athletes can be followed not only in their native language, but directly in dozens of languages by their fans all over the world.
4. Linum and Decoherence: Creating videos with AI
Systems such as Midjourney or Stable Diffusion can already produce impressive images that can hardly be distinguished from photos. But there are still a few systems that can also create videos using algorithms. The start-up Linum wants to close this gap. “Thanks to us, anyone can translate an idea into a film,” said founder Sahil Chopra, who is building the company with his brother Manu.
The start-up Decoherence, which was founded by former Amazon employees Rishi Bhuta and Will Stith, is pursuing the same goal. While Chopra declined to comment on the number of current users, Bhuta said his company already has more than 180 paying customers.
5. Baselit and Tennr: Analyze internal data
Large corporations have the resources to meaningfully evaluate and optimize their business. Small businesses often cannot afford this. Start-ups like Tennr and Baselit want to fill this gap. “A restaurant can, for example, evaluate what the five most popular products are on Saturdays,” said Baselit founder Sahil Singla.
Tennr wants to build an AI assistant right away, which can be fed with as much information as possible. These should not only be internal documents, but also public sources such as news articles, websites or LinkedIn profiles. This means that the system can be used by companies for their internal evaluations, but also by external experts who want to analyze a company or an entire industry using public sources.
More: How ChatGPT and Sam Altman will impact your life