Model training is similar to the creation of the cgi for the movie. Both happen before anyone consumes the output, and represent the up front cost for the producer.
Both a movie and a language model can cost tens or hundreds of dollars to produce.
In both cases additional infrastructure is needed for efficient usage: movie theaters or streaming platforms for movies, and data centers with the GPUs for LLMs. This is also upfront (capex) costs.
At consumption time, the movie requires some additional resources, per viewing, whether it's a movie theater or streaming. Likewise, an llm consumes some resources at inference time. These are opex. In both cases, the marginal cost for inference/consumption is quite low.
It’s almost like HN is a great platform for the POSSE model!
Awesome share thanks for the link. Will send to a family member who is looking to gain viewership with their writing - they usually post on medium I think.
We had a dog who would pull watermelon rinds out of the compost pile to eat. We gave her nice bones, but it's not enough. Nothing is enough. All is food and food is all.
How can you know that? Dogs were domesticated on being fed scraps to start with and came from wolves which are also natural scavengers and eat all sorts of nasty (to us) things. Ive seen many dogs happily gobble up insects, even stink bugs. Dogs don't even know what their food is made out of 95% of the time these days so im not sure anybody can claim they simply don't like eating insect food without some kind of study to back it up.
I was being a little facetious. Yes they probably "like" processed insect food designed for them, but for the average dog, I'm still betting on the steak.
Wolves scavenge opportunistically, but they are first apex predators. Their primary food drive is to hunt in packs for large game and gorge. Dogs are not so far removed.
reply