IOKA AI

At IOKA AI, we believe in the power of technology to enhance human potential. We're committed to a "human-in-the-loop" approach, exploring AI applications that benefit workplace processes—including areas like public sector education—while actively partnering with individuals and companies using these models to innovate with purpose. Our goal is to is to ensure that AI augments, not replaces, how we work, explore, and pursue our passions.

See our Knowledge Base & AI Glossary

Features

Artificial Intelligence

Neural networks, data classification and predictive modelling are nothing new, they have been at the core of Machine Learning for decades, but until 'ChatGPT' the best we could hope for was the predictive text you see on your smartphone messaging App - and hopefully an accurate weather report! However Vector Embedding has brought context and meaning as we engage with LLMs (Large Language Models) through Artificial Assistants, having virtual conversations (see Character AI) or converting text to image, text to moving image or text to music or audio.


The Models  2018 - 2024

The GPT (Generative Pre-trained Transformer) - first introduced by OpenAi in 2018, was then followed by a series of next step versions - until in November of 2022, they dropped 'ChatGPT' on us and changed everything! In July of that same year MidJourney went silently into Open Beta.

But the race in Generative AI really ramped up in 2023 when Anthropic released Claude and Alphabet Inc. introduced us to ‘Bard’ early in 2023 (later renamed Gemini) - that same year came ‘Mistral AI’ and Cohere released Coral. Meta released 'Llama' in April of 2024. Only Apple remained as the last sleeping giant - expectations of an iOS + AI device integration were realised when announced at their WWDC24 but to be honest, many expect more from them in 2025.

One of the more significant releases in this space came in December 2024 when Amazon announced 'Nova' as they stepped into the foundation model race.


The next steps

These foundation models continue to get smarter and throughout 2024 we have seen massive leaps towards ‘Artificial General Intelligence’ (see our predicted countdown clock). We have also started to experience a shift towards AI Agents that utilise many different models and 2025 shall see more Multi-Agent Systems (MAS) as organisations adopt or build enterprise automation solutions to find efficiencies and new processes. 2025 will also see attempts at implementing regulation with elements of the EU AIAct coming into effect seeking to set standards of governance and monitoring of AI Systems developers and providers - reducing the potential risks and impact of GenAi on our human rights, our data, our ways of working and protection over our creativity.