mbesto

And people who don't understand the difference between capex and opex are making uneducated claims. It's not basic math.

Running an inference data center is a mix of variable and fixed costs. The fixed costs are currently in the billions of billions of dollars for pretty much any investment in this space. Many of those fixed costs have (currently) unknown refresh cycles. So, unless you have access to the financial books of these companies it's currently just speculation whether inference is profitable.

> because one of the main issues with AI content is the ability to generate an overwhelming amount of it.

So then let's focus on that, and not whether it's generated by AI. Yeesh you people are hard to please.

That's what the parent commented on and that was my response.

  The final episode of The Late Show With Stephen Colbert will air on Thursday, May 21, 2026, at 11:35 p.m. ET/PT on CBS.Final Week Broadcast DetailsThe Finale Date: Thursday, May 21, 2026.Air Time: 11:35 p.m. ET/PT.Where to Watch: Broadcasts live on the CBS Television Network and streams live on Paramount+ for Premium subscribers (available on-demand the next day for ad-supported tiers).Finale Guests: CBS is keeping the final episode's lineup completely under wraps, though the preceding days featured major appearances from Jon Stewart, Steven Spielberg, David Byrne, and Bruce Springsteen.Context of the FarewellThe 11th season finale marks the conclusion of Colbert's decade-long run as host, following a controversial cancellation announcement by Paramount Global. While CBS publicly attributes the ending to a financial decision amidst a shifting late-night market, media analysts and former host David Letterman have heavily criticized the move. Many view the cancellation as a corporate effort to avoid friction with the Trump administration during critical regulatory mergers. Following Colbert's departure, CBS will retire The Late Show franchise completely and hand the time slot over to syndicated programming.Watch Stephen Colbert confirm his final broadcast date on Late Night with Seth Meyers:31sStephen Colbert confirms his final show date and reveals ...Late Night with Seth MeyersYouTube• Jan 28, 2026Are you looking for information on how to attend the remaining tapings, or would you like to know more about what Byron Allen show is replacing his time slot?

Could be worse - could be lead acid and weigh 2x as much and you only get half the Ah.

> You can see for yourself as it's AGPL and I assume you looked at the code

Can you look at any codebase and tell me it's written by some of the best engineers and it's not trivial?

For example - you won't get kicked out for leaving trash all of the ground but you will absolutely be shunned and shamed by everyone around you for doing so. That notion simply doesn't scale to a place like the US with 350M people with varying cultures, values, etc. because the social contracts are simply all over the place and inconsistent.

This IS the complaint.

Assuming it ever does...first, GPL is hardly enforced and second, I feel like there is going to be enough money (e.g. Anthropic's own code it uses for the harness) that pushes back against it being problematic. We'll see.

So these two things are squarely at odds with eachother...meaning, I don't know any PE acquirers who are actively terminating deals because the target acquisition's code is generating by an LLM even if the lawyers try to get a rep about it in the purchase agreement.

For the record, I still have yet to have an M&A lawyer explain to me unilaterally that AI generated code is an infringement...hence the question "who owns the code Claude Code writes" is still open.

I work in M&A. Nearly every lawyer, accountant, investor, and software business owner thinks their code is solely valuable and a trade secret. I find it hilarious and try to be as diplomatic as possible about why it's not. They also willfully will give their client list to a potential acquirer but get super cagey they moment a third party provider asks for their code to be scanned.

This argument easily gets shut down when I asked why, Twitch, a $1B business didn't crater to their competition when their full codebase was leaked.

The big improvement then was a carbon plate. Adidas (and others) followed suit. The subsequent improvements since then have been marginal but the margins are thin at that level. In this case the big advancement has been the weight of the shoe.

EDIT: Also it's worth noting these shoes are $500 retail. Adidas will for sure get a boost in sales from this, but there's definitely competition in the $200~$300 marathon running shoe space that won't solely draw everyone to Adidas)

"In the aerobic exercise domain up to ~100% of maximal oxygen uptake (VO2max), CHO is the dominant fuel, as CHO-based oxidative metabolism can be activated quickly, provide all of the fuel at high aerobic power outputs (> 85-90% VO2max) and is a more efficient fuel (kcal/L O2 used) when compared to fat."

https://www.gssiweb.org/sports-science-exchange/article/regu...

I think it has more to do with the fact that many of the products built for defense are never actually used against adversaries in their useful life. Just look at our nuclear weapon stockpile.

Palantir on the other hand is an invisible weapon. They could be reading my comment right now and identifying me with sentiment "adversarial" for all I know. What implications that has on my daily life is innumerable...and I'm a US citizen!

VidStudio invokes FFmpeg — a free multimedia framework — to handle certain video and audio processing operations. FFmpeg is licensed under the GNU General Public License v2 (or later).

The FFmpeg WebAssembly binary is not hosted or redistributed by VidStudio. Your browser fetches it directly from the public npm mirror at cdn.jsdelivr.net the first time you use a feature that requires it.

FFmpeg source code is available from ffmpegwasm/ffmpeg.wasm (the WebAssembly port) and git.ffmpeg.org/ffmpeg.git (upstream FFmpeg). The full text of the GPL v2 license is available at gnu.org/licenses/gpl-2.0.

You don't set up k8s because your current load can't be handled, you do for future growth. Sometimes that growth doesn't pan out and now you're left with a complex infrastructure that is expensive to maintain and not getting any of the benefit.