A16z invests in text-to-image AI generator Ideogram

25 Aug 2023

Image: © besjunior/Stock.adobe.com

Ideogram AI has received $16.5m in seed financing to develop its ‘state-of-the-art AI tools’, but it enters a competitive and sometimes controversial market.

Andreessen Horowitz (A16z) has backed a new text-to-image start-up trying to make its mark in a sector dominated by AI models such as DALL-E and Midjourney.

A16z and Index Ventures co-led a $16.5m seed financing round into Ideogram, according to the start-up. Other backers include AIX Ventures, Golden Ventures and Two Small Fish Ventures, along with various industry experts.

Ideogram claims to be developing state-of-the-art AI tools that can make creative expression “more accessible, fun, and efficient”. The company is developing its own AI model that can generate images based on simple text descriptions.

“We are pushing the limits of what’s possible with AI, with a focus on creativity and a high standard for trust and safety,” Ideogram said.

The start-up claims to have a team of “world-renowned AI experts”, who have worked on various other AI projects such as Imagen, Google’s own text-to-image system.

While it may be hard to stand out in such as flooded market, VentureBeat claims Ideogram’s AI model sports reliable text generation, which could give it an edge for generating logos or similar creations.

Text-to-image highs and lows

Before the rise of OpenAI’s ChatGPT and similar AI chatbots, text-to-image generators were the hot topic in the AI sector.

One of the biggest names in this sector was OpenAI’s DALL-E – which was released in 2021 – and its successor DALL-E 2 which was unveiled in April 2022.

A publicly available AI model called DALL-E Mini also gained a lot of attention last year, being able to reproduce the results of OpenAI’s text-to-image model with a much smaller architecture.

Various companies have shown support for these types AI models, such as Shutterstock which launched its own AI image generator earlier this year. Others have taken issue with how some of these models operate, however.

At the start of 2023, Getty Images sued Stability AI for allegedly stealing copyrighted content for its AI art creator, Stable Diffusion. In 2022, Getty Images issued a ban on the upload and sale of AI-generated images on its platform.

The company said this ban was due to “open questions” surrounding the copyright of AI-generated images, along with uncertainty surrounding the data these AI models are trained on.

In September 2022, JumpStory co-founder Jonathan Løw told SiliconRepublic.com that there is a growing “legal minefield” around using AI-generated images for commercial purposes. He also said that the legal risk may fall on the end user if their commercially used image enters a copyright dispute.

The commercial use of AI-generated images suffered a legal hit recently, when a US court ruled that this type of art cannot be copyrighted. Judge Beryl Howell of a US district court said that human authorship is a “bedrock requirement of copyright”.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.

Leigh Mc Gowran is a journalist with Silicon Republic