Yeah you can sample the beat and put your own lyrics over it, it's been proven in court as fair use

To use the dialog someone has to 1) prove you didn't write it without evidence and 2) prove you didn't rewrite portions of it, also without evidence. That's a pretty high burden of proof for the person trying to copy something they could have just asked chatgpt to generate for them.

You only need power to train models. Once you train a perfect or near-perfect AGI model.... how much power do you need to run it? Survey says about 100w instantaneous which is about what a human generates. But it only needs 100w when doing an actual task, which is usually 3-20 seconds at a time, a couple times an hour. How much do you use chatgpt?
 
I suspect there will be a huge spike in power needs, but actually running the model will be less than 20w per person per day

You'd need to identify and prove beyond a reasonable doubt that
 
1. that particular game asset was AI generated, and
 
2. the asset had not been modified in any way after generation, since modifying the asset makes it an original work and thus copyrightable
 
There's an extremely thin slice of AI generated work, like chat responses, first pass image generation that doesn't have copyright, but also has zero or close to zero economic value. You don't get to pirate the entire game GTA VI simply because there's a single 2D unaltered sprite in a folder somewhere that was ai generated

I doubt investors will get $0.10 on the dollar for their investment. Buying internet forums has been a losing proposition for everyone. No collaboration is needed this IPO is signal enough that the investors want out and will unload their shares on anyone unlucky enough to buy this dumpster fire of no growth

That will likely be the top price before it gets delisted at the end of the week

Right but as an investor you're only going to invest if you predict additional growth to buoy the stock price, looking for a dividend, or take over of the company. This IPO is merely a parachute and it's going to tank the valuation

Like most IPOs these days, this is the investors trying to get their money out, not raising capital to grow the company. If they could have made more money via private investment to fund growth, they would have.

We share the market with 3-10 other suppliers and there's a finite number of customers. While your statement is true for commodities, knowledge workers typically don't have much of a direct impact on those kinds of end products.

Then the law will change. Countries that update their laws to allow LLMs to continue to function in their current state will see new datacenters built and those services will just run from there. Outsourcing an tiny crappy text based API to an international location is so little effort it's not even worth explaining how one would do it.
 
Besides, outside of a handful of megacorps, it'll be totally impossible to track. You can run a pretty dang good LLM on an 8 year old computer with 8GB ram (which is most of them) that runs at human speed.
 
The government isn't going to allow all that money to flow outside of the country. I expect to see verbiage like "directly derivative works" included in the laws in the future; i.e. you can't make Star Wars 19 with Seinfeld, but you'll be able to make Space Movie with Comedians from NYC without paying royalties.
 
There's way, way too much money to be made with AI compared to what the entertainment industry produces in any given year.

We're not hiring more but we expect to do additional volume (20%) with the same number of workers over the next couple of years. Just being able to summarize PDFs is a huge time saver by itself. The efficiency gains are obvious, a significant amount of backoffice tasks can either be fully or partially automated. You'll always need a human to handle override situations, but instead of a team of five accountants processing monthly expense reports, you might only need two, a lead, and a backup in case the lead quits.
 
Just today apple announced an audio tagging feature to summarize audio in video clips for their video editing software. Pretty tiny incremental upgrade, but 30 seconds here and there add up to an hour a week and if you have 40 video editors on staff doing different projects, that's an entire headcount you can drop without working anyone harder. Audio summarizing might actually save more like 5-10 minutes a day depending on workflow and how close the editor is to the source content. Could be over an hour a day if the editor isn't fluent in english and is really struggling to understand what they're editing.

You know you have problems if you're taking research money and are having trouble maintaining your tax exempt status. Things have to be super fucked up to have your tax exempt status in question.

Bard LLM (or any LLM, really) is going to be 1000% better than whatever pre-LLM technology they were working on for their voice activated assistant technology. That went from "voice assistant is an impossible nut to crack" to "college students build/train better technology in a weekend as a homework assignment" in less than two years.

Bitcoin was created in 2008 and it's 2024 and still to date hasn't found a useful purpose
 
Compare to AI where we're actively integrating it into our daily processes to make them more efficient
 
Blockchain is dead dead dead. The only people still trading it are those trying to unload it at a minimal loss, or those day trading the volatility.

My mom has been trialing Hulu for about three months ahead of her cable tv subscription coming up for renewal. She dropped cable TV in january. She is in her 70s. I don't know anyone who has cable at this point. If people in their 70s are switching to streaming then cable TV has a Very Real Problem.