exodust

Censoring the statue of David is not old news! Unless you go back to the 19th century:

  "The practice of covering Michelangelo's "David" with a fig leaf, particularly a detachable one, was common during the Victorian era, specifically in the mid-19th century, due to Queen Victoria's discomfort with the statue's nudity."

So we have a newly tightened trajectory of censorship. Grok was promoted as an AI that strives to "maximize truth and objectivity". We've gone from that to effectively placing fig leaves over statues.

Meanwhile the linked article fails to clarify the outcome, which apparently was a "rogue prompt change" by a new employee, Musk wasn't involved. The prompt was reversed and "the system was functioning as intended." It's completely understandable there might be a volume of anti-[personX] material on the internet, put there as propaganda, that might serve as weighted reasoning for an LLM. This is nothing new, unless you're aiming for cheap shots.

Only juvenile prompting and reporting finds value in "what nasty thing did the LLM say about Musk."

What's more interesting is the censorship on the input side. I wanted Grok to analyse a wood engraving from 1861, Gustave Doré's "Harpies in the Forest of Suicides". Wouldn't do it. Grok's content policy filter refused to accept the upload because..."boobies". So I pixelated the wood-engraved breasts and re-uploaded. This time it worked. [1]

What about Michelangelo's statue of David? Denied. Instead of pixelating, I cut out David's genitals and placed them on his leg, leaving a genital-shaped hole in the groin area. [1] Bingo. Image accepted. Genital location matters. Now wasn't this more interesting than "Musk-Bad-Man-Says-Grok"?

https://imgur.com/a/pIdBJXm

I like the perpetual license, no AI, and customisable 3D models and animations. This last feature hopefully opens up potential for making creatively synchronised graphics such as animating the expression - say the mouth shape on a 3D face in response to modulation. I wonder if the animation needs to be a fixed amount of frames or length?

> When Indian users use writing suggestions from an AI model...

Let's press pause and note the deliberate decision in using AI to help you with your "favorite food & holidays".

A better remedy to the problem invented by this article, is to recommend against using any AI for writing about your personal experiences, values and preferences.

As an Australian man, my favourite food is [shrimp on the barbie mate] Thai noodles. You see? If my AI model was drenched in Australiana culture, it will come up with stereotypes I don't subscribe to.

  Me: can dynamic content such as inventory feeds be included in wiki pages?

  *AI Assistant is typing*

  AI: Hmm, I couldn't find an answer to that. Can you rephrase your question or give me a bit more detail?

This is why I can't stand the idea of conversing with AI bots just to "browse" a company wiki. I mean how big are company wikis? Not big enough that simply browsing it yourself using regular content browsing or keyword searching can't surface what you need quickly and accurately.

And $790+ annually and still can't remove the "powered by Perfect wiki" logo! It takes $2390 before you're unsticking that sucker!

I don't remember pushback against CSS, but I do recall it was common for websites to dabble in CSS for styling fonts but still use table layouts. That's not pushback, just practicality.

If you wanted a solid predictable grid, you used table layouts. Particularly for tight and complex layouts. It wasn't uncommon for the navigation section to be a table layout even if other parts weren't. "Pixel perfect" was the name of the game. There emerged some pushback against the elimination of pixel-perfect design, not CSS.

Around 2000, remember spacer elements?

  <spacer width="332" height="1" type="block"></spacer>

They never caught on. Pixel GIFs were more common and predictable.

Until flexbox went mainstream in 2014, CSS float layouts were an exercise in balancing spinning plates. Empty divs didn't behave the same as divs containing content, and it sucked. We'll never get those hours back, but at least we were paid.

The idea that early CSS gave us wonderful "zen garden" layouts doesn't align with my experience. Then again, I was making information media sites not fun arty sites.

> needing summarization

Before we get too excited with disparaging those seeking summaries, it's common for people of all levels to want summary information. It doesn't mean they want everything summarized or are bad people.

I'm not particularly interested in "tariffs, what are they good for, what's the history and examples good or bad"... so I asked for a summary from grok. It gave me a decent summary. Concise and structured. I asked a few follow-ups, then went on with my life knowing a little more than nothing about tariffs. A win for summarized information.

Your example doesn't appear to contain a reply from grok, only a question.

True. I see that too. It's a good addition to community notes. It can correctly evaluate "partially true" posts and those lacking details, so it's great at spotting cherry-picked information.

And here we go. SEOs scrambling to get "rewarded" in AI-driven search results. The article takes a swing at uncovering the secrets of "how to get your ideas included in more LLM responses."

In this case the LLM was spot on to flatten not reward! In no way is her idea an "original and innovative insight" as claimed. Something about a "multilingual SEO matrix", and the "coining" of that term. The LLM has seen it and noticed nothing special or original. Good LLM, have a biscuit!

Her idea is a breezy mash of existing concepts around multilingual SEO. Just because you're proud of your late-night SEO matrix chart, doesn't mean it should be elevated to epic status without time and consensus from others to evaluate usefulness.

> LLMs are incapable of attributing an original concept to its creator

This just isn't true. It might be for something you coined last week by mashing together existing concepts in the most unremarkable way imaginable. In time, valuable original concepts are indeed tied to their origins that AI knows about.

You've evaluated a tiny sample of restaurant websites and extrapolated to make claims about the overwhelming majority - in the millions, across the globe.

"Most are generic crap" doesn't mean restaurants aim for that benchmark when they decide to get a website.

I'm not sure if you're refuting the point I was making, which I'll clarify. "Restaurant website" could be a stand-in for any basic small business website. The claim was that AI threatens small web dev agencies who make small business websites. I don't think it will, as millions of small businesses want something better than "generic crap" or cookie-cut AI copy paste; AND we've had site-building services, social media pages, and template-driven approaches for a long time.

Still use Google for quick generic lookups, spell-checks/definitions, shopping stuff, products, or things I know will return a good result.

Grok is great for finding details and background info about recent news, and of course it's great for deep-diving on general knowledge topics.

I also use Grok for quick coding help. I prefer to use AI for help with isolated coding aspects such as functions and methods, as a conversational reference manual. I'm not ready to sit there pretending to be the "pilot" while AI takes over my code!

For the record, I do not like Google's AI generated results it spams at me when I search for things. I want AI when I choose to use AI, not when I don't choose it. Google needs a way to switch that off on the web (without being logged in).

3-5 prompts doesn't get you a professional restaurant website.

The key word is professional. A good restaurant website begins with taking good photos of the premises and the food. AI won't come around to your business and take professional photos.

There's a lot of bits and pieces to a website for bookings, content management, menu updates, etc.

HTML templates and themes have been around for a long time. AI can basically spit out those templates and themes, which is great. But there's still a lot to do before you get to www.fancy-dining.com.

They are not stealing! Indiana Jones is already out there in pop culture.

We shouldn't complain about AI holding a mirror up to our world and noting "you guys love Indiana Jones a lot. Here's a picture inspired by his appearance, based on your generic prompt that I'm guessing is a nod to the franchise."

The AI is a step ahead of your unsubtle attempts to "catch it stealing".

The image of Indiana Jones is not "ready for market" when it emerges from your prompt. Just like Googling "Indy with whip", the images that emerge are not a commercial opportunity for you.

When you make multi-billion dollar movies with iconic characters, expect AI to know what they look like and send them your way if your prompt is painfully obvious in its intent.

If you're handing out red cards for repetition, aren't you guilty of the same infringement? You repeated the exact same prompt, which is itself an obvious bait for AI to draw the pop-culture lookalike.

I'm surprised by the comments here. When your prompt is so mind-numbingly simple, an obvious "dare" for AI to be a naughty little copyright infringer, can you blame it for dishing up the "big mac" you asked for while sitting in a fancy restaurant?

Don't want Indiana Jones? Don't ask for Indiana Jones!

The prompt is what's lacking creativity. The AI knows it, and produced the obvious popular result for a bland prompt. Want something more unique? Write better prompts.

Spread on toast with goats cheese, sauteed mushroom & fresh chilli; poached eggs on top of all that. Brew coffee and there's work-from-home breakfast sorted.

> AI is the best programming tutor you can wish for...you can always ask for clarification or examples when something isn't clear

Yep, AI can teach beginners the fundamentals with endless patience and examples to suit everyone's style and goals. Walking you through concepts and giving you simulated encouragement as you progress. Scary stuff, but that's how it is.

But... as we know, it doesn't always provide the best solution. Or it gets muddled. When you point out its mistakes, it apologises, recognises the mistake and explains why its a mistake. It's reasoning is incredible, but it still makes mistakes. This could be very risky for production code.

Related anecdote... I needed Photoshop help recently for horizontally offsetting a vignette effect. Surprisingly not easy. The built-in vignette filter can't be applied to a new blank layer, and is always centred on the image. AI suggested making it manually but I didn't want to do that, as I like the built-in vignette better. AI's next solution involved several complicated steps using channel isolation and weird selection masking etc. No thanks. Then my own brain sparked a better idea... simply increase the canvas size temporarily, apply the vignette, then crop back to the original size. Job done. I told AI about my solution and it was gushing with praise about how brilliant my solution was compared to its own. Moral: never stop trusting your own brain.

> "writing down exactly what he wants" is just coding!

Agreed. The more exact and clear your instructions are, the closer to programming it is. Presumably the non-technical person has an application where they care about things like performance, scalability, compatibility and all those things coders sweat over.

Fair enough. I mentioned "initial fun". I enjoyed it to a point, but recall it abruptly ends and seemed unfinished, unresolved. The berating was cute at first then gets old.

Compared with a game like Portal, with similar self-referential quirkiness. Even had its own meme "the cake is a lie". That was a resolved game. Stanley on the other hand... I won't hammer the criticism because I'm glad games like this exist.

Never understood the appeal of stanley parable or beginners guide. After the initial fun wears off, the constant abuse and sarcasm from the narrator is abrasive and noisy. Not as clever as they're hoping. The Beginners Guide even worse. Empty test levels strung together with tedious narration.

The developer's depressed? That's the least interesting thing to learn about anyone. Depressed people promoting their depression sounds disingenuous.

Not to be all negative, for walking simulators with no interaction but nice dreamy atmosphere, try Liminalcore. Relaxing and calming, great on an ultrawide OLED. Lots of vast shadow areas. Huge scale architecture. Bump the FOV setting up a little from default. No cheap gimmicks, just a restrained dream world with subtle hint of something lurking in the shadows. In terms of artistic walking games, this is one of the good ones.

https://store.steampowered.com/app/3107900/Liminalcore/

> Prove the claimed fraud.

This is a forum not a court of law. You're hung up on "fraud" and lost sight of the discussion.

Sometimes "fraud" is, for the time being, billowing black smoke. There's a small chance it's a harmless smoke machine, not a fire. Some will cling to that hope, but many others will yell fire.

The process with DOGE is unfolding, so hang tight! Surely you agree "efficiency" would be a good outcome. They're aiming for a tight ship not a cash-haemorrhaging behemoth.

Let Musk work on it for a bit. Heckle later if he doesn't get results. I'll join you in heckling, but for now it's great what they're doing. Shocking that Teslas are burning.

You originally replied to someone who asked "How much fraud and abuse...", so they weren't tethered to only "fraud", they were talking about the generally alarming and wasteful spending of the public's money.

I'm not an American and I'm not expressing my likes or dislikes in this matter. I'm interested in seeing a healthy America. I'm interested in the details, drama, action, science and politics which trickles down or across to the rest of us in the world.

You mention morality, but then say my "likes" are nonsense. You've even pushed me into the "rant" and "hate" zone! I'm trying to have a civil discussion about curious things. Insinuating I'm ranting and hating doesn't help.

Maybe we differ on the morality point. In my mind, anything immoral is a problem worth solving.

The law sometimes trails behind what the community needs. In my country Australia, the law sometimes doesn't protect the community from harm (bail laws, police strip-search powers, weak foreign investment review boards, lack of industry oversight allowing criminal gangs to infiltrate construction industry including major government contracts - yes these are real and recent examples.) We don't want the law to fail. We like "the law", and because we like it, when it does fail we need to face it, name it, and deal with it.

> Show that the money was spent in a way different than proscribed by law

When people find loopholes in the law to obtain grants, do you defend their actions because technically it didn't break the law?

DOGE is apparently exposing a system of dubious and questionable spending at scale. If "wasteful" is an understatement, don't be surprised when taxpayers use "fraud" to describe the decisions made to blow their money.

Even if a court finds that funding that "drag show in Ecuador" was lawful, doesn't take the sting out. It doesn't take the fact away that someone benefited who probably shouldn't have, relative to the government's broader budget purposes. It might enrage people even more that it was a systemic abuse of grant money.

> recent (regrettable) turn

It's only regrettable through the lens of partisan politics. DOGE is undeniably a good concept.

You will notice governments around the world copying the method. That is, launching initiatives to audit everything, cut wasteful spending, expose corrupt spending, and trim bureaucratic inefficiency via external oversight. It would be regrettable not to undertake the task.

What? The parent article is a SpaceX mission post. Meanwhile I note you have contributed 2 posts in this topic and both of them are about Trump. Do you have any thoughts on the actual event that took place? Or is your needle stuck in Trump's groove?

Again, this thread's parent article is a SpaceX mission launch post stating a brief outline and details. And the top comment links to some rubbish tech-wanker article.

Again, the topic is "crew returning to Earth." You, Berger, and whoever posted the arstechnica link have brought politics to the party.

I liked how they had a drone circling the recovery activity at the splashdown site. All live-streamed for our informed entertainment. SpaceX is undeniably doing a great job, although the starship explosions in our atmosphere I'm not a fan of, I look forward to that not happening.

> "they were re-scheduled"

You're over-correcting and understating. "Nothing to see here" is not how we describe 8 days turning into 9 months.

Don't be angry about hype. Everyone has access to what the astronauts say themselves. In this case, their remarks have been ambiguous, probably because they don't want to be drawn into a shit-storm. They didn't want to be "victims" in any way, so they were not "stranded". But their Boeing shuttle did suffer problems, and that isn't nothing, nor is it a "scheduling issue".

End of the day, it was great to see the parachutes open. Health complications can arise for astronauts stuck in space that long. Their return was overdue to the point of being concerning for most people. Fair?

The topic is "crew returns to Earth" not arstechnica's routine political spin.

Does arstechnica mention the dolphins even once? Nope. How about Trump? Six times! Meanwhile, sibling comment applauds the writer for his "knowledge of space topics".