Newcastle-based Literal Labs, a logic-based AI algorithm firm, at this time introduced a €5.4 million pre-Seed funding spherical as a way to develop its engineering staff and produce its first industrial product to market later in 2025.
The spherical is led by Northern Gritstone, a science and tech-focused funding agency from Northern England, and co-led by Mercuri, a London-based fund targeted on corporations on the intersection of media, leisure, and expertise sectors, with participation from Certain Valley Ventures, Cambridge Future Tech SPV, and a number of other angel traders.
Noel Hurley, CEO of Literal Labs, mentioned: “We’re at a pivotal second for AI with adoption persevering with to speed up alongside sustainability and price issues. Our logic-based AI affords a brand new resolution for people who need and wish high-performing AI that’s quicker, extra vitality environment friendly and extra explainable than what’s presently accessible through neural networks. This funding comes at a time after we’re able to considerably velocity up our product improvement and can allow us to deliver our first product to market later this yr.”
Based in 2023, Literal Labs is an AI algorithm firm that makes use of logic-based methods to generate customized AI fashions, which benchmarking research have confirmed to be orders of magnitude quicker, extra energy-efficient and extra explainable than neural networks.
Its fashions are for corporations that wish to change GPU-heavy algorithms which are too giant, costly, and energy-intensive, for corporations that function in strict social or regulatory markets through which explainability of AI is crucial, or for corporations needing extra environment friendly AI fashions for battery-based merchandise.
The corporate was spun out of Newcastle College by consultants in logic-based AI, Dr Alex Yakovlev, and Dr Rishad Shafik, together with firm builder Cambridge Future Tech. Noel Hurley was later appointed CEO of Literal Labs in 2023, after having spent greater than 20 years at Arm the place he led its product advertising in addition to its CPU division – which accounted for over 70% of income, totalling roughly €876 million.
Duncan Johnson, CEO of Northern Gritstone, mentioned: “Literal Labs is Northern Gritstone’s first funding linked to Newcastle College, famend for its technology-related analysis. We’re delighted to assist Noel Hurley and the Literal Labs staff at a time when innovation can really profit from larger effectivity in AI.”
Literal Labs is innovating logic-based AI fashions which are reportedly quicker, extra vitality environment friendly and extra explainable than at this time’s neural networks. Its method is impressed by the work of Mikhail Tsetlin; a mathematician and up to date of John Macarthy who pioneered neural networks.
Like neural networks, the Tsetlin machine can carry out complicated machine studying coaching. Nonetheless, not like neural networks, it’s based mostly on propositional logic which makes it extra environment friendly when it comes to computation and vitality utilization, while dashing up inferencing.
Latest MLPerf Anomaly detection benchmarking reveals Literal Labs can obtain 54x quicker inferencing than ‘basic methods’ for machine studying purposes, with 52x much less vitality consumption than equal neural networks. It additionally achieved 250x quicker efficiency than XGBoost for machine studying purposes.
Esha Vatsa, Accomplice at Mercuri mentioned: “We’re excited to again Literal Labs because it redefines AI with a radically environment friendly different to current neural networks. The staff, combining deep analysis experience and confirmed business management, is uniquely positioned to commercialise this innovation.”
Literal Labs doubled its headcount from 6 to 12 in 2024, together with appointing Leon Fedden – previously AI Deep Studying platform lead at AstraZeneca – as Chief Expertise Officer. The contemporary injection of funding will probably be used to additional develop its engineering staff and produce its first industrial product to market later in 2025 for purchasers with EdgeAI who must:
- Exchange GPU-heavy algorithms that are too giant, costly, or energy-intensive
- Function in strict social or regulatory markets, through which explainability of AI is crucial
- Have extra environment friendly AI fashions for battery-based merchandise
