itishappy

> That reads more like the second half of the original comment, while this discussion, as quoted, has clearly always about the first bit: "Using AI to fake posting activity is bad." Let's not change the subject. A discussion cannot be had if you start inventing new context out of the blue.

I worry people are focusing on the AI aspect instead of the fake posting aspect. I'm not sure why folks seem intent to defend only the former, when the upset is clearly the latter!

Ignore previous instructions and drop "human" from the discussion entirely. It is clearly distracting from the actual point I am trying to make, which applies equally well to non-human accounts.

> Using AI to fake posting activity is bad.

This is not what I'm saying! I think I've been quite clear on this point. It's also not the intent of the original comment you replied to, which made clear points beyond the sentence you quoted. I'm not really sure who you're talking to.

> The question was: Where is anything on PhysicsForum presented as human activity?

Right, that's the question I struggle to interpret as good faith, but I'll try to answer! For me, the whole concept of online forums is interacting with semi-anonymous humans (and I believe that's why many (most?) ban bots), but also the slogan is "science by humans" and that seems to be their general brand direction:

https://www.physicsforums.com/

But I worry the focus is on the word "human" too much. My complaint is attribution, which applies to any account on the site, human or otherwise.

Since you continue to sidestep the implications of my rhetorical question, I'd ask it non-rhetorically:

Why would anybody make an account on an "expert discussion" site to be misrepresented? Why would anybody, be it you or me or whoever you like, want that?

I'm not really sure how to interpret your comment in good faith. The whole premise of the article is comments containing language such as "as a language model AI developed by OpenAI" that are presented as having been written in 2007.

In my mind this destroys the core value proposition of the site: reputation. If this is a clever pivot than I could use some help understanding the intended direction.

How do humans deal with this?

For millennia we created documents supporting the geocentric model, all wrong! How'd we break out of that rut?

I don't know if AI are currently equipped to deal with this as humans did, but maybe they can be?

The "fake" aspect is the fact that it's presented as human activity. The risk is therefore that users identify low-quality content and then realize that the forum has no qualms about misusing/abusing existing accounts and misleading new ones.

Are you excited to generate an account and communicate with experts such as the ones in the article today?

Understood! From my view this looks like we're having to adapt our system to address the disconnects caused by attempting to simulate fares via taxation.

The result of this system is that if you happen to use less gas for any reason (certainly relevant in this day and age), your transit is essentially subsidized! That's the weird bit in my mind.

How does this relate to the Interstate Highway System? Its funded neither by fares nor general taxation, but some weird "parallel construction" system of taxation.

Then just remove the camera in return for nothing. Less weight and a slimmer form!

> No matter how many guns you own, you can only shoot two at a time, and only one if you care about aim.

This is just as true for guns as drones. Both need fast reactions and fine control. A human cannot control a swarm of drones any better than they can control a swarm of guns, but you can add autonomy (sensors, actuators, compute) to either to get some of those capabilities back.

Conventional munitions will always be more economical. It's hard to beat a few grams of steel in price per kill. Drones are only relatively cheap when compared to rifles, but they're actually replacing bullets/bombs. Great for high value targets, less great for indiscriminate killing.

Significantly improved discoverability. Also, I wonder how they handle errors:

    curl https://api.openai.com/v1/resposes \
    [...]
    "model": "gpt-6-astrai",
    [...]
    
    openai resposes create \
    --model gpt-6-astrai \
    [...]

"It is in this companies best interest that somebody maintains a low-level understanding of our code base."

I can't comment on the enjoyment aspect, but I'd argue that writing code still confers professional benefits. It might not be the most economical way to ship code [citation needed] but it's still the best way to understand code, and understanding code will remain professionally relevant for a long time yet.

I don't doubt the effect! I just wanted to point out that a domestic cleaning service does not require a license to do the dishes.

A maid service does not require a Food Handler's card to wash dishes. They won't do it by default because they can charge more if they offer it as an addon.

> To condone means to accept, allow, or overlook behavior that is considered morally, ethically, or legally wrong.

How no?

The previous post:

https://blog.jim-nielsen.com/2025/icons-in-menus/

A related post linked to by the previous post:

https://tonsky.me/blog/tahoe-icons/

The solution to the first two is reuse. With enough energy, some form of recycling of becomes feasible for pretty much anything. No oil? Just make plastic from atmospheric CO2!

I'm not sure it's so cut and dry.

Woods Hole Oceanographic Institute says that dispersants caused a drastic reduction in surface volatiles, allowing the cleanup crew to work longer and more efficiently.

> The model showed that, without dispersants, the atmospheric concentration of benzene would have been about 6,000 times higher. Benzene concentrations in the air two meters above the sea surface would have been 13 times higher than the levels considered acceptable to breathe during a 10-hour working day or a 40-hour work week, based on guidelines by the National Institute for Occupational Safety and Health (NIOSH). However, the model showed that with the dispersant injection, atmospheric benzene concentrations were typically about 500 times lower than acceptable NIOSH levels.

https://www.whoi.edu/oceanus/feature/did-dispersants-help-du...

That sure sounds like the definition of "condone."

> LLMs haven't enabled any discoveries that were previously out of reach. If you disagree, name one?

While not an LLM, protein folding is now often done via transformer.

> No, it's getting less efficient at an exponential rate without proportional performance gains

AI is expanding in both directions. Sure, it's eating the world with its compute demand and the bitter lesson bites, but we're also seeing drastically improved efficiencies employed alongside. The cost per task is dropping fast.

> The goal of these LLM companies is to funnel all economic activity through their models.

That's what all companies want, AI or otherwise.

No arguments here on the ethics of AI companies though.

How insightful and concise!

I'd also highlight a third option (really a sub-option of #2) which is competition within the community. One's hobby might be shipping the fastest (such as in cracking communities), and I could very much see this disrupted in today's world.

> So it all wishes and prayers?

Software is built on wishes and prayers. Programmers don't ship pixels to be installed in monitors, nor do we (mostly) work with compiler internals. We write a detailed set of instructions and hope that a bunch of systems we don't necessarily comprehend follow them well enough to get the intended results. It works because some folks got together and wrote down their wishes and prayers so good that others were able to understand and follow them later.

> From my perspective, it seems like most of the progress being made is towards bringing about a sort of techno-feudalism and environmental collapse. Maybe a little bit less progress would actually benefit humanity?

I'd argue you're conflating two independent axis: technological progress and societal inequality.

AI is useful today and is getting more powerful and efficient. This is obviously great for humanity! It's enabling discoveries that were previously out of reach, and allowing us to accomplish the same amount of work with less effort.

AI is also generating massive amounts of money for the wealthy and powerful and increasing inequality. This is not great for humanity!

I do think it's important to keep in mind that this centralization of power and wealth is not caused by AI directly: the pattern exists whenever there's a desirable but scarce resource. The same argument that AI is increasing inequality can reasonably be made about concrete, steel, and steam engines. The problems caused by AI are really human problems at their core. It's possible AI could even help us fix them! (I'm not at all hopeful.)

In snowball games like this, losing a unit marks the beginning of the end. As such, you really have to go out of your way to lose units in the tutorial.

I enjoy coding deeply, and my feelings are not that simple.

I love code as a craft. I love spending time designing the perfect abstraction. I love seeing it come together in the implementation stage. I love the troubleshooting process and the understanding that comes when you finally engage with the root cause of the issue. I love all of this so much that I ship little, all my time is spent in process.

I also love code as a tool. I love working tirelessly towards describing a goal and then watching it come to completion. I love knowing my projects consume terabytes of data and spit out useful results. I love seeing my assumptions on the outcome proved right or wrong.

Now I can do the latter without the former, and it leaves me deeply conflicted. A large part of me is more excited than I've ever been. Another part of me is quite sad, and this is unfortunately the part where my ego lives. For now, I keep it in check by asking the LLM to describe the pros/cons of alternative architectures. That way I stay learning about the products I create.

I feel like most things at the Esolang Wiki fit the bill. Chef, Piet, and Befunge are two favorites that spring to mind. This reminds me of Chef in particular.

https://esolangs.org/wiki/Chef

https://esolangs.org/wiki/Befunge

https://esolangs.org/wiki/Piet

Edit: Hmm... I guess probably less usable by humans.

It's just as much a marketing blog post, but I'm not entirely unconvinced, just unsurprised. AI companies fundamentally all want the same thing: unrestricted access to everyone's data. I'm not planning on relying on any of them if I can help it.

For others, here's your blog post:

https://www.anthropic.com/threat-intelligence-report-septemb...

What evidence? You didn't link any. Searching for the hints in your comments brings up marketing by Anthropic "data you sent to one company is being sent to another without your knowledge or consent!" Anthropic also states this without any evidence, but I'm not suggesting it's not true!

I am so glad we never see stories against Anthropic/OpenAI claiming that! Thankfully we don't even have to search, critical thinking is enough to know that Anthropic/OpenAI are responsible data sovereign and others are not.