if you want to take OpenAI’s own research into account

No thank you.

OlympicArena validation set (text-only)

“Our extensive evaluations reveal that even advanced models like GPT-4o only achieve a 39.97% overall accuracy (28.67% for mathematics and 29.71% for physics)”

• The OlympicArena analysis that you cited.

Source?

Probably defacing PBS?

Though I think that may have kindof been the FBI’s idea rather than Anonymous’. (It took place after Sabu was compromised by the FBI and I think it was kindof Sabu that started… was it LulzSec that that PBS hack was done under?)

Fuckin’ FBI.

(Also, don’t know why you’re getting downvotes.)

Unlike this year when LLMs are more of a huge scam.

You can’t really and make a profit. You pay more in electricity than you get in crypto.

…unless someone else is (unknowingly) paying for the electricity.

(Of course, when the price of crypto takes an upturn, sometimes it might get profitable again. And I’d imagine there are people mining it even when the price is low banking on the idea that it’ll spike again and they can sell it.)

No joke. I’m ashamed to say I have had to endure Weblogic in the past. God was that time a massive clusterfuck.

The company I worked for decided to use two particular separate products (frameworks, specifically; ATG and Endeca, even more specifically) to use in tandem in a rewrite of the company’s main e-commerce application. Between when we signed on the dotted line and when we actually started implementing things, Oracle acquired the companies behind both products in question.

The company should have cut their losses, run away screaming, and started evaluating other options. That’s not what happened. Instead, they doubed-down and also adopted several other Oracle products (Weblogic and Oracle Linux on (shudder) Exalogic servers) because that’s, of course, what Oracle recommended to use with the two products in question. The company also contracted with Oracle-licensed “service integration” companies that made everything somehow even worse.

And the e-commerce site rewrite absolutely crashed and burned in the most gloriously painful way possible. They ended up throwing away tens of millions of dollars and multiple years on it.

When the e-commerce site rewrite did happen, it was many years later and used basically only FOSS technologies. I guess at least they learned their lesson. Until the upper management turns over again.

The sooner the crypto bubble bursts, the fewer victims there will be of fraud like this.

You think the OP image is AI? How do you figure? (Or maybe my brain just put too much emphasas on the word “without” and you aren’t saying the OP image is AI.)

So Wario, then? Maybe that makes Android Waluigi.

Is the human traffic cone someone I should recognize?

Did he really post this? Where? TruthSocial?

Removed by mod

A buck says her next tatoo is a swastika.

And was born 370 million years too late to sit under the shade of prototaxites.

It does depend what you consider “the whole series.”

Does Tingle RPG count? Does Link’s Crossbow Training?

Other questions that probably wouldn’t matter for best or worst game in the series, but would for a complete ranking might be whether remakes/remasters count as separate games or not and if not, do we judge on the basis of the remaster or the original? (Either way, all the remasters are inferior to the originals. Maybe that’s my most radical, non-left-right position. That and Breath of the Wild was better than Tears of the Kingdom.)

Shhh. We don’t talk about that one.

It’s like eating fingernails. 🤮

The worst thing about the 2011 Green Lantern movie is how it killed any chance of more Green Lantern movies for so long.

(Ok, to be fair, it wasn’t just the 2011 movie. Also the pandemic. And DC’s complete ineptitude at movies.)

The big U-haul loves his baby U-hauls so much he carries them everywhere with him.

In this thread: Cryptobros downvoting every realistic take on cryptocurrency for being “bearish” and “FUD”.