Clover-Edition VS aid_adventure_vulnerability_report

AI Dungeon Clover Edition isn't exactly the best in terms of output quality in comparison to most alternatives (mainly because it just hasn't been updated in quite a while; its best models still use GPT-Neo 2.7B), but it's the most similar alternative to Classic AI Dungeon, and plays more like a game than current AI Dungeon, or any of these other alternatives.

KoboldAI with GPT-J 6B (or a more powerful open-source model) is probably the best free alternative. You can either run the AI model locally, or use the KoboldAI GPT-J 6B Google Colab. Dreamily also exists as a free option, but I'm honestly not sure I'd recommend it over KoboldAI + GPT-J 6B (or even AI Dungeon, for that matter). If you want something that plays more similarly to the CYOA-style of AI Dungeon, try AI Dungeon Clover Edition. It's not exactly the best free option in terms of output quality, but it's pretty much the closest one in terms of gameplay to AI Dungeon.

The best you can get are some pre-dragon modified versions like Clover Edition if you want AID specifically.

There are also a ton of branches of the original version of AI Dungeon, but Clover Edition is likely the most notable, and arguably the best out of them. It also has a link to a Google Colab version, in case you can't run it locally. It's worth mentioning, though, that it's not really quite as good, in terms of output quality, as other options, mainly due to it using GPT-Neo 2.7B. It also hasn't been updated in quite a while. However, as far as AID alternatives go, it's one of the alternatives that are closest to the AI Dungeon experience, if that's what you're looking for in an alternative.

There are plenty of free alternatives. Write With Transformer exists, is free, and has a few AI models to choose from. You can try the base GPT-J 6B model on EleutherAI's website, or through the KoboldAI Google Colab. Clover Edition exists, and has multiple AI models to choose from. GPT-Neo Dungeon exists, and uses GPT-Neo, hence the name. Open CYOAI and AI Dungeon 2 Unleashed also exist. GodAI exists as well, and uses GPT-2. KoboldAI is a good frontend for locally running AI models, and it has a subreddit at r/KoboldAI. Dreamily exists, and has a mobile app. However, they require quite a bit of personal information to make an account; only make an account if you trust them with said info. Their privacy policy also, at one point, actually admitted to monitoring private content. There is also a filter in place, though I'm unsure what the filter disallows (aside from mentioning Xi Jinping). Also, HoloAI, Hyperwrite, ShortlyAI, and InferKit have fairly abusable free trials. HoloAI uses GPT-J 6B, InferKit uses Megatron-11B, and Hyperwrite and ShortlyAI both use GPT-3 (although, they both also use OpenAI's incredibly broad filter; quite a lot of stuff is disallowed).

Clover Edition exists, and has multiple AI models to choose from.

Getting an older version of the game isn't a great idea. If you mean an older version of the app, the filter would be there anyways, since the filter is done on the server-side, and the AI would be the same as it is now. The original version of AI Dungeon uses the Classic AI, which I doubt many people are willing to settle for. However, there are plenty of forks of the original version of AID, and some AID clones, such as Clover Edition, GPT-Neo Dungeon, Open CYOAI, and AI Dungeon 2 Unleashed. As was mentioned already, NovelAI also has a text adventure mode that plays near identically to AI Dungeon, and the text adventure module does kinda capture the feel of AI Dungeon. Honestly, all it takes to recreate AI Dungeon on alternatives is to add "> You"/"> You say" at the start of your inputs. Literally all AI Dungeon's Do and Say modes did was add "> You" and "> You say", respectively, to the beginning of your input. Also, I think you're underestimating how many people actually did use AI Dungeon to write serious stories.

NovelAI uses GPT-Neo 2.7B and GPT-J 6B, with subscription tiers of $10/$15/$25 per month. HoloAI uses GPT-J 6B. It has a free trial, and subscription tiers of $5/$8 per month. Hyperwrite offers 1500 outputs for free, and uses GPT-3. After the free trial, a subscription is required. Though, some content (to be more specific, NSFW content) is disallowed. ShortlyAI uses GPT-3, and offers a free trial. After the free trial, a subscription is required. Though, similarly to Hyperwrite, NSFW content is disallowed. Write With Transformer exists, is free, and has a few AI models to choose from. You can try the base GPT-J 6B model on EleutherAI's website, or through the KoboldAI Google Colab. Endless Visual Novel, though it isn't yet released, will be going into closed alpha this month (you can currently sign up for the closed alpha, if it interests you). A subscription (the subscription prices haven't been specified yet) will be required to use the AI and create stories. It uses the GPT-3 Davinci model (the same model that AI Dungeon's Dragon model is based on) for text generation, and also has AI-generated imagery and music to go along with it. As the name implies, it plays in a visual novel format. Depending on what you use AI Dungeon for, it could genuinely end up outclassing AI Dungeon when it releases. There are also some more alternatives worth mentioning, though they all require some setup to get working, and most run locally, which requires more computing power than what the average person has. While they are quite good, the fact that they're aren't exactly user friendly might be a deal breaker for you. Anyhow, Clover Edition exists, and has multiple AI models to choose from. GPT-Neo Dungeon exists, and uses GPT-Neo, hence the name. Open CYOAI and AI Dungeon 2 Unleashed also exist. GodAI exists as well, and uses GPT-2. KoboldAI is a good frontend for locally running AI models, and it has a subreddit at r/KoboldAI.

HoloAI uses GPT-J 6B. It has a free trial, and subscription tiers of $5/$8 per month. Hyperwrite offers 1500 outputs for free, and uses GPT-3. After the free trial, a subscription is required. Though, some content (mostly NSFW content) is disallowed. ShortlyAI uses GPT-3, and offers a free trial. After the free trial, a subscription is required. Though, similarly to Hyperwrite, some content is disallowed. InferKit has a free trial, uses the Megatron 11B model. A subscription is required after the free trial. Write With Transformer exists, is free, and has a few AI models to choose from. You can try the base GPT-J 6B model on EleutherAI's website, or through the KoboldAI Google Colab. Clover Edition exists, and has multiple AI models to choose from. GPT-Neo Dungeon exists, and uses GPT-Neo, hence the name. Open CYOAI and AI Dungeon 2 Unleashed also exist. GodAI exists as well, and uses GPT-2. KoboldAI is a good frontend for locally running AI models, and it has a subreddit at r/KoboldAI. Dreamily exists, and has a mobile app. However, they require quite a bit of personal information to make an account; only make an account if you trust them with said info. Their privacy policy also, at one point, actually admitted to monitoring private content, as well as to collecting quite a lot of personal info and sharing said info with third parties. There is also a filter in place, though I'm unsure what the filter disallows (aside from mentioning Xi Jinping).

The main privacy issues were because of Latitude, at one point, manually reviewing stories that triggered the censorship filter, and because of the data breach back in April of 2021. Latitude said a while back that they'd no longer be monitoring private stories, and although they could theoretically read your private content, they have absolutely no reason to do so, and it'd just be a waste of their time. The vulnerability that allowed the 2021 data breach to occur was also supposedly fixed.

Additionally, there was a data breach that occured about 2 weeks prior to the filter controversy, and that the AI Dungeon community only found out about during the controversy, apparently allowing all private content, except for private content from the first 11 days of AI Dungeon 2's existence, to be accessed. Latitude failed to disclose the breach for months, which people were understandably upset about. It wasn't until 5 months after the breach occurred that Latitude finally disclosed the breach.

To learn that our stories had been breached through a third-party is ridiculous. I wonder if you would have just ignored it if AetherDevSevOps' vulnerability report never dropped.

My pleasure. Here is the official data leak and the entire story of how the penetration tester approached it, including his prior security warnings he tried to give to Latitude that they completely ignored: https://github.com/AetherDevSecOps/aid_adventure_vulnerability_report

Latitude also announced the content policy, which disallows a lot of content. It also turned out that Latitude trained the AI on pedophilic content, among other content that their content policy disallows, which a lot of people saw as being pretty hypocritical. Although, Latitude did say in the past month or so that they removed the questionable content from the finetuning data. Latitude also officially disclosed the data breach, although they did so about 5 months late. Also, during the past month or so, Latitude finally broke their silence. Nick Walton (Latitude's CEO), as well as Latitude's new CPO, have been fairly active on this subreddit during the past few weeks, and for once, Latitude is actually being somewhat transparent.

Feedback form is not working at all since the data breach (they disabled every form to prevent further leak) and every official social networking account except Twitter disowned or went silent

The Incident is actually several things that all happened around the same time, the two big ones being the implementation of The Filter, and a white-hat hacker publicly sharing a data breach he found that allowed him to view all unpublished, account-level stories dating all the way back from 2017. Much of the The Controversy discussion is around The Filter, and while it was poorly implemented (and the follow-ups to it make it even worse, especially the auto-ban "feature"), what got me to leave was the Data Breach.

Uh... what? The filter is broken, and the fact that phrases like "You thrust your sword at the infantry soldier", "You analyze the young lady's wounds", and "The boy equips his breastplate, in preparation for their sparring session", are all guaranteed filter triggers, regardless of the context (you can test them yourself, by the way) basically confirms that the filter is a regex filter that works as described here. The data breach did happen, though, for some reason, the creator of that report edited out a lot of it a while back; before it being edited, there was some pretty definitive proof in the form of a bunch of inputs that were made during the time of the vulnerability. You could've actually found inputs that you made, if you played during the time that the data was collected from. The AI was trained on pedophilic content, among other questionable stuff. You can even download the original AID training data to confirm this, yourself. While this technically isn't "wrong" for Latitude to do, since they have the right to censor whatever they want, there is a lot more than just pedophilia that's disallowed. I will say, though, there are some misconceptions and outdated info regarding the filter. People still seem to think horses and numbers below 18 can trigger the filter; they actually used to, but they don't anymore. When it comes to numbers, you now have to say "[number below 18] year(s) old" to be able to trigger the filter. Watermelons being filtered is also a common misconception, caused by a guy who got flagged for asking for 4 watermelons. He got flagged for 4 (this was when numbers on their own were filtered), not for watermelons.

Additional shady acts There’s been a major security breach that’s still yet to receive any official statement.