What’s FOSS-AI? A model everyone can download and use for free? Or in the OSS spirit that everything need to be open and without discrimination of use, aka OSS training data corpus and no AUP attached?
Or you mean the inference engine running those models?
Everything which is not BigTech. Preferably FOSS, at least not BigTech, just alternatives to for example OpenAI.
So you’re including free models like freeware, not FOSS only, by non big tech.
Your choice of models will be quite limited as the compute resource and training corpus needed to make a viable base model isn’t anyone can do.
There are VERY FEW fully open LLMs. Most are the equivalent of source-available in licensing and at best, they’re only partially open source because they provide you with the pretrained model.
To be fully open source they need to publish both the model and the training data. The importance is being “fully reproducible” in order to make the model trustworthy.
In that vein there’s at least one project that’s turning out great so far:
If a layman may ask, what are folks even using AI/LLMs for mostly? Aside from playing around with some for 10-15 mins out of simple curiosity, I don’t have a practical use for platforms like ChatGPT. I’m just wondering what the average tech enthusiast uses these for, outside of academia.
I teach language. I get paid for my time in front of students, not the time it takes to prepare their lessons and the materials. I use AI to quickly reference grammar rules, to fabricate example dialogs in specific scenarios to practice, and to suggest activities to do in class to practice the target grammar. I never do exactly as it says, just take it as kind of a source of suggestions for me to build from.
That sounds like a time saver for sure. I imagine that some of those elements (grammar rules) are widely available everywhere, while others (practice dialogues, activity suggestions focused on the use of language) would require a fairly specific training model.
Well, LLMs are quite literally trained on language, so asking it to simulate a conversation between a hotel clerk and a guest who is upset that they can’t find the hair dryer is pretty much what it’s best at doing.
You can even build the dialogs with students. Have them introduce a scenario for the LLM to manufacture, then have the students suggest variables to apply, such as the clerk being hungry and in a bad mood while the guest is actually drunk after returning from a club in order to see how the language changes, then have the students act it out for laughs.
A friend of mine and I have gotten used to using it during our conversations. We do fast fact-checking or find a good first opinion regarding silly topics. We often find it faster than digging through search-engine results and interpreting scattered information. We have used it for thought experiments, intuitive or ELI5 explanations of topics that we don’t really know about, finding peer-reviewed sources for whatever it is that we’re interested in, or asking questions that operationalizing into effective search engine prompts would be harder than asking with natural language. We always always ask for citations and links, so that we can discard hallucinations.
Thanks for sharing! I’m probably too set in my ways to ever utilize AI for things like this. I never use virtual assistants like Alexa or Google either, as I like to vet and interpret the source of information myself. Having the citations would be handy, but ultimately I’d want to read them myself so the IA/VA just becomes an added step.
I’m just convinced all of y’all asking about this are in a huge circle jerk that never ends, but refuses to understand how it all works.
A model is a model. It’s a simplified way of narrowing down thresholds of confidences. It’s a pretty basic sorting algorithm that runs super fast on accelerated hardware.
You people seem to think it’s like fucking magic that steals your soul.
Don’t send information over the wire, and you’re golden. Learn how it works, and stop asking dumb questions like this is all brand new, PLEASE.
There is a difference between a general scare about the AI buzzword and legitimate distrust in online services which are closely connected to american spying institutions (regardless if they are ai or not)
If my calories tracker app would apoint a (former) NSA official on their board, I would be looking for alternatives too. This is not about AI, this is about a company with huge sets of private data being closely interconnected with american spy institutions.
Sad that you don’t seem to be able to distinguished between legitimate security questions and badly informed hypes/scares ass soon as a buzzword like AI occurs
