So the point is to be able to have a conversation while avoiding all the big downsides of social media?

Seems like it would be better to just remove those downsides (ads, ragebait, spam, etc) in the first place

Great idea - are you sure however that all of the "AI Generated" examples are actually AI? What about stuff made by good old fashion human + photoshop ? Obviously this would still be "generated" or not real but I think it is a somewhat important distinction

I would guess if you have it export multiple files from the same source in one go it already does but could be wrong.

This post is how meta engineers just recently submitted a patch with the ability to avoid starting a new process for every output encoding and so they can share the decoding step. Maybe that also includes sharing the motion estimation step, but I would be careful making such assumptions, FFMPEG has a lot of low hanging optimization work that hasn't been done just because someone hasn't done it yet.

Or under clocking and under volting for even better performance to price/power/longevity ratios

For a single rack, you really don’t have too many choices for power. You make a choice to provision and pay, I never had anyone check how much of that I used and give me money back. Maybe things have changed though.

No doubt. Especially for GPU inference at scale. We overclock/overvolt for training and tune way down for inference.

Or just restore from backup across the board. Assuming they do their backups well this shouldn't be too hard (especially since its currently in Read Only mode which means no new updates)

Agreed - and its a huge step up from their previous naming schemes. That stuff was confusing as hell

I see your point. I do find Anthropic's approach more clean though particularly when you add in mini and nano. That makes 5 models priced differently. Some share the same core name, others don't: gpt 5 nano, gpt 5 mini, gpt 5.1, gpt 5.2, gpt 5.4. And we are not even talking about thinking budget.

But generally: These are not consumer facing products and I agree that someone who uses the API should be able to figure out the price point of different models.

I dont think the average power user needs to understand DNS. Knowing just that it can be changed (and can fix things or break things and whatnot) is probably already plenty.

Connecting to SSH seems like something a "power user" should be able to learn but not necessarily know already (probably more likely they know what a VPN is)

Very cool

Seems like this could also be used by call centers to realtime adjust their accents. Text is obviously easier to analyze (no realtime required) but I imagine that audio is not that hard to process real time.

Calling for home internet support and getting the person on the other end (in a US Southern or Boston accent) asking you to "do the needfull" could be pretty entertaining :-D

Why bother with accents when you can replace the call support workers alltogether with AI? Isn't that why all AI companies have gorillions in valuation?

KeePass ftw

Just make sure you have backups