true, most, maybe 95% will take an easy path

I paid for chatgpt, it just a different beast to feed...

venice.ai now is on shitcoin route

duck duck go ai offers access to multiple models as well privacy and at no cost

the bonus with deepsake is high performance on low budget GPU, this allows anyone to download it and connect to their own rig or cloud computing retaining control over the data with a great privacy, self hosted AI is the way to go. it's open source like llama is but more efficient. the true power of LLM comes with the agents - hopefully deepsake will accelerate AI agents experience and make it even more affordable

most of the new iPhone users don't need all the power coming with the CPU

most drivers don't exceed 100 mph

most AI users use it for basic stuff and pendrive stored large language models without the Internet will be enough

This LLM is loading in the browser and stays there, can work offline afterwards and is not sending data anywhere, it's all local and private

For image generation, just an example but it's doable this way too. I'm quite hopeful

https://huggingface.co/spaces/webml-community/janus-pro-webgpu