- N E W S I N C O N T E X T -

Using large language models (LLMs) trained on biological diversity at scale, a California startup revealed this week that has demonstrated the first successful precision editing of the human genome with a programmable gene editor designed with GenAI.

The development is aimed at improving the way scientists fight illness and disease and is part of a wider effort to build AI technologies that can improve medical care.

Just like OpenAI’s ChatGPT learns to generate language by analyzing books and lots of other written texts, Berkeley-based Profluent’s technology created new gene editors after analyzing enormous amounts of biological data, including microscopic mechanisms that scientists already use to edit human DNA. These gene editors are based on Nobel Prize-winning methods involving biological mechanisms called CRISPR. Technology based on CRISPR is providing a way of altering genes that cause hereditary conditions.

Gene editing has the potential to solve fundamental challenges in human health as well as in agriculture and biotechnology. However, “CRISPR-based gene editors derived from microbes, while powerful, often show significant functional tradeoffs when ported into non-native environments, such as human cells,” says an abstract of a paper Profluent is expected to present next month at the annual meeting of the American Society of Gene and Cell Therapy. “Artificial intelligence (AI) enabled design provides a powerful alternative with potential to bypass evolutionary constraints and generate editors with optimal properties.”

Profluents said that it is “open sourcing” this editor, called OpenCRISPR-1, so that individuals, academic labs and companies to experiment with the technology for free.

Read on to get the key takeaways from this story and the week's most important technology news impacting business.

It is no wonder that utilities are having trouble keeping up with the surging demand for electricity to power AI data centers. A recent International Energy Agency (IEA) report forecasts that data centers’ total electricity consumption could reach more than 1,000 terawatt-hours(TWh) in 2026, demand roughly equivalent to the electricity consumption of Japan.

As AI gets incorporated into software programming in various sectors, demand is expected to spike even more. The study forecasts that search engine sites like Google could witness a tenfold increase in electricity demand when AI is fully implemented on its platform. When comparing the average electricity demand of a typical Google search (0.3 Wh of electricity) to OpenAI’s ChatGPT (2.9 Wh per request), and considering 9 billion searches daily, this would require almost 10 TWh of additional electricity in a year.

“Updated regulations and technological improvements, including on efficiency, will be crucial to moderate the surge in energy consumption from data centers,” says the report.

That is where Heimdall Power comes in. The Norwegian scale-up has developed an eight pound “magic ball”, a sensor the size of a bowling ball that once fastened to live power lines by drones, can increase existing power line capacity by 30% to 40%.

How is that possible? Power lines get hotter as flow increases. If they get too hot, they shut down. Without accurate sensors a power company sets limits on the flow of power on a seasonal basis rather than in response to real-time conditions, which can lead to a lot of unused capacity.

“Think about the temperature on the line as being the speed limit,” CEO Jørgen Festervoll, said in an interview with The Innovator. “Without accurate information utilities are driving without a speedometer so they drive conservatively. Our technology allows them to get as close to the speed limit as possible.”

That’s important because as the world moves away from fossil fuels more electricity is needed not just for AI but for clean energy. “Often these projects are told that the grid is full and to come back in seven to 10 years, the time to build a new line,” says Festervoll. “That is not good for the economy, and it is not good for the climate. Society is saying we want AI, we want Cloud services, we want electric vehicles, but you can’t have all of that and renewable energy unless there is enough capacity on the grid.”

- I N T E R V I E W O F T H E W E E K -

- S T A R T U P O F T H E W E E K -

Vaultspeed, a Belgian scale-up, helps corporates organize and centralize their data in one place in a fast and easy way so that they can fully leverage the power of artificial intelligence. Customers include banks, insurance companies, government, manufacturers, energy companies and makers of luxury goods.

“If you are a company like a bank, you can’t differentiate yourself on AI alone because another bank will be using the exact same AI provided by Big Tech,” says Vaultspeed CEO and Co-founder Piet De Windt. “The way to differentiate yourself is by being faster and better in bringing data together and organizing it better. If you do that you will have a competitive advantage because the broader the data set and the better structured, the better the AI will understand and be useful in specific use cases.”
Vaultspeed is one of 48 deep tech scale ups that were selected to join The EIC Scaling Club, a curated community that includes investors, corporate innovators and other industry stakeholders. (The Innovator’s Editor-in-Chief is on advisory board of The EIC Scaling Club, an arm of The European Innovation Council).

- N U M B E R O F T H E W E E K