What have you found to be an effective way to tell if you're chatting with a bot or a real person?
So I’ve just tried it with chatGPT, and it replied normally. I asked it why it wasn’t bothered by Cyrillic letters, and it answered this:
I am designed to understand and generate text in multiple languages, including those that use the Cyrillic alphabet like Russian, Bulgarian, and others. So, Cyrillic letters don't bother me. You can continue the conversation in either English or a language that uses the Cyrillic alphabet if you prefer.
So it really depends on the model, I guess. Don’t rely too much on this advice.
I'd say a similar amount of time, but I post and comment significantly more here. And I like kbin's vibe a lot more than I did reddit's for the last few years, so I don't even miss it. This place is nice :)
So I tried it on this BBC article (a current top story), and this /r/Hearthstone post. It did pretty well. I won't copy-paste the whole reply, but here are some excerpts:
The post you've shared describes a series of recent climate records related to temperature, ocean heat, and Antarctic sea ice, and their concerning implications. [...] The trustworthiness of this information depends on the credibility of the sources cited and the accuracy of the climate records mentioned. Given that these are attributed to credible individuals and institutions like the UN, they likely hold a high degree of trustworthiness. [...] A pattern of increasing intensity and frequency of extreme weather events is consistent with what scientists expect from climate change. [...] However, for a comprehensive analysis, it would be prudent to cross-check this information with authoritative climate research bodies, like the Intergovernmental Panel on Climate Change (IPCC), National Aeronautics and Space Administration (NASA), or National Oceanic and Atmospheric Administration (NOAA).
The post you've shared appears to be an announcement about an upcoming expansion for Hearthstone. [...] The new expansion, named "Legendary Titans and Keepers," seems to introduce some new gameplay elements, including the "Titan" keyword and "Forge" keyword. Assessing the trustworthiness of this information can be tricky without an official source. Ideally, the announcement should be verified on Blizzard Entertainment's official website or through their official social media channels. The details mentioned, such as the gameplay mechanics for the new Titan and Keeper cards, as well as the new Forge keyword, are specific and elaborate, which might lend some credibility to the post. [...] If this information came from an official announcement from Blizzard Entertainment or a reliable insider, it would be newsworthy content for audiences interested in Hearthstone or gaming in general.
So it guessed correctly in both cases and suggested where to fact-check the info to be sure.
Did you intend to paste or attach something? Your comment doesn't show anything on kbin besides that one sentence.
Half of the deleted […] things are chatGPT mentioning its 2021 knowledge cutoff and suggesting double-checking that info. It was mentioned in this case as well.
If it were an autoGPT with internet access, I think these would prompt an automated online lookup to fact-check it.
It didn’t really take 250 years though, early emissions were almost negligible. Most of it started like 60 years ago. You’re right that we’re not stopping it anytime soon, but the effective timelines are shorter than centuries.
Also, what’s your reasoning/source on a 10 bn “absolute” cap? It might be a cap while using modern farming, technologies and logistics, but it’s not absolute by any means. You mention beaming energy from space, then why not mention Eucomenopolis concepts that allow for trillions of people to inhabit Earth? :) Or simply once you have fusion, you can have vertical farms and Arcologies that can sustain a much larger population.
The issue isn’t that it’s impossible, rather that we’re not gonna develop any of this tech before humanity faces existential problems in many parts of the world.
Also, it’s weird that you got from “this temperature variance is minimal” and “this average is on the low side of comfortable” to “let’s get rid of 30% of population then”. oO
Let me put it this way. The difference in average global temperature between the last ice age 100,000 years ago and pre-industrial earth around 100 years ago is just 3.5C. The expected temperature rise due to recent climate change is about +3C.
A lot is going to happen, and much sooner than in 50 years.
With current trends, it looks like we’re heading towards severe climate destabilization, much more common extreme weather events, some parts of the world becoming uninhabitable for humans, lots of mass extinction events for many species, including those that humanity currently relies upon, and probable global famine.
The fact that it’s not a complete extinction doesn’t make it fine. Sure, Earth as a planet will be fine. But the civilisation has some really hard challenges coming up, and it’s currently not prepared for them.
Half of the deleted […] things are chatGPT mentioning its 2021 knowledge cutoff and suggesting double-checking that info. It was mentioned in this case as well.
If it were an autoGPT with internet access, I think these would prompt an automated online lookup to fact-check it.
Air temperature graph starts from 1940, that’s 83 years. Enough to gauge trends, since industrialization and copious CO2 emissions in particular are a pretty new thing.
But here’s some data starting from the year 0, in case you’re interested.
I just found it kinda funny that the rule is actually wrong irl since yes is more common across the board, yet when formulated as a question the answer to it is no :)
Remember that since this is a planet-wide average, it includes places like the North Pole and Antarctica. Or just look at the graphs — it’s a pretty visual demonstration of how extremely abnormal recent temperature changes are.
Publications like these lower reddit’s future valuations though. I’m not sure whether a two-day uptick in users is worth that. My guess would be that it’s not.
Reposting my comment from another similar thread to show that this is easily fixable, and you should be wary of any non-reputable news source anyway.
So I was curious how current LLMs might handle this with proper instructions, so I asked chatGPT this: “What can you tell me about this Reddit post? Would you write a news article about this? Analyze the trustworthiness of this information:” and pasted the text from the post. Here’s a part of its reply:
This Reddit post appears to be discussing updates in the context of […] Hearthstone and World of Warcraft. However, there are several factual inaccuracies which raises questions about the trustworthiness of the post:
- Hearthstone was first released in 2014, not 1994. This error may be a typo or a misunderstanding by the user.
- Glorbo does not appear to be a known character in either the Hearthstone or World of Warcraft universe […]
- Chen Stormstout is a known character from the Warcraft universe, but the claim about him being the end boss of a new version of Karazhan doesn't match with any known information […]
- Dragonflight is not known as a video game developer or associated term in the context of Hearthstone or World of Warcraft […]
Given these points, it seems this post might be either a joke, speculation, or misinformation. […] So I'd recommend taking this information with a grain of caution and verify it from reliable sources.
[…] I would caution against using the information in this post for a news article due to the factual inaccuracies and lack of verifiable sources. Good journalism relies on accuracy, fact-checking, and use of reliable sources. This Reddit post contains several claims that contradict established knowledge, and without further verification from reliable, official sources (like game developers or official news releases), it would not be advisable to use this information in a news article.
So it’s not even an issue with current models, just bad setup. An autoGPT with several fact-checking questions added in can easily filter this stuff.
Can any headline that ends in a question mark be answered by the word no?
No.
Here’s a quote from that wiki page you linked to:
A 2018 study of 2,585 articles in four academic journals in the field of ecology similarly found that very few titles were posed as questions at all, with 1.82 percent being wh-questions and 2.15 percent being yes/no questions. Of the yes/no questions, 44 percent were answered "yes", 34 percent "maybe", and only 22 percent were answered "no".
In 2015, a study of 26,000 articles from 13 news sites on the World Wide Web, conducted by a data scientist and published on his blog, found that the majority (54 percent) were yes/no questions, which divided into 20 percent "yes" answers, 17 percent "no" answers and 16 percent whose answers he could not determine.
I mean, chatGPT with its knowledge cutoff and no internet connection figured it out. See my comment below, I asked it and posted its response.
The guys who run that news website just didn’t include any checks in their algorithm. It doesn’t seem like an LLM problem at this point. A properly set up AutoGPT with an ability to look stuff up online would have no problem sorting though and fact-checking posts to decide which ones to use for an article.
I was curious how current LLMs might handle this with proper instructions, so I asked chatGPT this: “What can you tell me about this Reddit post? Would you write a news article about this? Analyze the trustworthiness of this information:” and pasted the text from the post. Here’s a part of its reply:
This Reddit post appears to be discussing updates in the context of […] Hearthstone and World of Warcraft. However, there are several factual inaccuracies which raises questions about the trustworthiness of the post:
- Hearthstone was first released in 2014, not 1994. This error may be a typo or a misunderstanding by the user.
- Glorbo does not appear to be a known character in either the Hearthstone or World of Warcraft universe […]
- Chen Stormstout is a known character from the Warcraft universe, but the claim about him being the end boss of a new version of Karazhan doesn't match with any known information […]
- Dragonflight is not known as a video game developer or associated term in the context of Hearthstone or World of Warcraft […]
Given these points, it seems this post might be either a joke, speculation, or misinformation. […] So I'd recommend taking this information with a grain of caution and verify it from reliable sources.
[…] I would caution against using the information in this post for a news article due to the factual inaccuracies and lack of verifiable sources. Good journalism relies on accuracy, fact-checking, and use of reliable sources. This Reddit post contains several claims that contradict established knowledge, and without further verification from reliable, official sources (like game developers or official news releases), it would not be advisable to use this information in a news article.
So it’s not even an issue with current models, just bad setup. An autoGPT with several fact-checking questions added in can easily filter this stuff.
That’s a weird argument, since cycling is the most efficient mode of transport. Even ignoring all the health improvements/lower emissions/etc., it still easily outpaces everything else in terms of environmental benefits due to its efficiency. It’s just not a good way to heat your home.
Sustained human pedal-powered energy output is about 500W for world-class cyclists and around 100–200W for average people. Your body also produces 100W of heat energy by simply existing, and that can rise to about 500W when exercising.
So the output range we’re looking at here is something like 300–1000Wh per hour depending on your fitness level and exercise intensity. 1 kWh costs 10–30 cents around the world, I think.
You’re gonna spend much more on extra food to fuel your pedalling than you’ll ever be able to save on heating bills :)
If Polish troops enter, for example, Lviv or other Ukrainian territories, they will stay there. And they will stay there forever.
Holy mother of projection.
I went a bit further and prefaced your question with this: “Analyse this question. What might have been intended by its author? What is asked here? Then, answer that question.”
And omfg:
Edit: tried it again, and it’s not consistent. It also goes for the letter T or the number 7: “…the runic symbol ᛐ might resemble the number 7 in our modern numerical system, albeit a slightly stylized or slanted version of it.”