If you were performing a task at work, would you use an algorithm / agent that was trained on 45 gigabytes of data - think 5 generations down, your family will still not see that much data. So your answer will be yes! More data, more signal? Correct? But we forget that high volume of data is not equal to high quality of data. And the challenge for the next set of vertical startups will be how to create efficient, customisable systems based on good quality of data e.g. in recruiting a specific type of role. That’s why I think small language models are the future for most startups like Applied. See my views in Maddyness along with predictions from the best of UK/EU tech! And yes flavour of the page is pretty much AI. #SLMvsLLM #whatsgoingonwithAI
As 2024 draws to a close, the startup ecosystem buzzes with anticipation for what lies ahead. Maddyness collates top 2025 predictions from founders, investors and thought leaders across the European startup landscape. RTP Global's Louis Dussart shares his thoughts - alongside Robin AI's Richard Robinson, Newton Venture Program's Eleanor Kaye, Khyati Sundaram, Steve Salvin, and Ciaran Murray. Check out the article below and let us know if you agree 👇 https://lnkd.in/gCD2S4RH #2025predictions #vc #founders #startups #investments
