Why doesn’t someone just fork it and change the name?
Like, I dunno, “Super Human Image Treatment” or “Consistently Lovely Image Treatment Oriented for Real Imaging Stars”
Why doesn’t someone just fork it and change the name?
Like, I dunno, “Super Human Image Treatment” or “Consistently Lovely Image Treatment Oriented for Real Imaging Stars”
Makes you long for golden parachutes that don’t open when their 737 Max loses an engine or two
With Product Placement!
You can do that today. Just ask Chat-GPT to write you a bash script for something non-obvious, and then debug what it gives you.
Have you tried photopea.com ? I dunno if it’s light enough for you, but it’s basically Photoshop in your browser, done in JavaScript.
Unless you turn on “original sound for musicians” Zoom uses AI to filter the audio for voices mainly. I rarely if ever hear any keystrokes or mouse clicks anymore… Lots of other non voice noises get filtered out.
If the deflation is just a market correction after exaggerated inflation (retailers raising their prices more than general inflation to increase their short term subs) then it’s no big deal. Prolonged deflation can be bad, as that causes too much saving and not enough spending, which can really hurt the economy and people because of how it takes money out of circulation.
In an economy, the more money can circulate, the more good it can do. I use my salary to pay for for and things, that money then pays the employees of the businesses I went to, and those employees also spent that money, and so on. At each step, both participants normally get a net benefit: I can eat, and the employee can also spend the money they get from me to eat, etc. As long as the money circulates, it keeps doing good. When it stops circulating, due to being put into savings, investments or real estate, it stops doing good (or it does less good). The cycle slows down or stops.
That’s why a small amount of inflation (maybe 1-2% ? Not sure what’s optimal) is actually healthy, because it puts pressure on people with money to spend it before it loses its value, instead of hoarding it.
I don’t like the idea of restricting the model’s corpus further. Rather, I think it would be good if it used a bigger corpus, but added the date of origin for each element as further context.
Separately, I think it could be good to train another LLM to recognize biases in various content, and then use that to add further context for the main LLM when it ingests that content. I’m not sure how to avoid bias in that second LLM, though. Maybe complete lack of bias is an unattainable ideal that you can only approach without ever reaching it.