Pertin-ant 🐜

Old post but relevant nevertheless. All AI is as smart as the data it is trained on; and this data is limited while the training is centralised Pertin-ant is the first peer to peer intelligence network that is aiming to decentralise the creation, collation and categorisation of data, and rewarding folks such as ourselves while it is at it. More on Pertin-ant at https://pertinant.net #AI #AIandblockchain #decentralisation #AItraining

The market opportunity for AI training data is huge. It is also a market that lays the foundation for #generativeAI The global AI training dataset market size was valued at USD 2.39 billion in 2023 and is projected to grow from USD 2.92 billion in 2024 to USD 17.04 billion by 2032, thus exhibiting a CAGR of 24.7%. Pertin-ant, in its early stages, will incentivize the global internet populace to create conversational data, in textual form. Incidentally text training datasets lead the training data set market with a majority of the share, approximated to be 33–35% of the total market (the other segments being image, video and audio). Text training data sets are forecasted to be the majority of the market, compared to the other segments, in 2032 as well #generativeAI #AI #AItrainingdata #trainingdata Read on...

The potential of #AGI in industry and in our daily lives, is well — transformative !!. LLMs (and LLMs are just one example of AGI tooling) are cavernous in their requirements of data; what is however happening is (and is borne out by hard facts) that we are heading towards finitism in the datasets that are needed to explore, and capitalize on, AGI’s infinite possibilities. Drum roll… Here comes Pertin-ant. Pertin-ant is the only of its kind, internet, web3 and blockchain project that enables the global internet populace , to monetize their knowledge and daily interactions, accelerating the development of AGI. Do read on.. #AI #AGI #AItrainingdata #AItraining #LLMs #LLM https://lnkd.in/gQNA_VZ8

Data is the foundation for #AI. While we are boisterous about AI as a transformational technology, a super important fact is that there simply isn't isn’t enough data to train all the world’s AI models in the future ; and the data that is there is not diverse and representative enough. We are steadily in a cul-de-sac here First of many blog posts from us at Pertin-ant 🐜, shedding a loud light on the data insufficiency at #AI #AI #Blockchain #web3 #medium #transformation

LLMs like #ChatGPT and #Llama require anywhere between 5-11 trillion words as a training data set. These training data sets are public data. Conversely, a mere 1-2% of the global internet populace creates this data/content. What are the limitations hence with these data sets and as an extension with #LLMs and #AGI