That’s just asking me my favorite adventure game, which is not a question I’ve ever found very interesting or answerable.
1 Like
My immediate response to ‘most famous adventure game’ was Zork of course. It was a commercial brand name even if obscure to some generations. I might venture that there are people who may conflate Zork with Colossal Cave and not realize they are different games. I played CC first in grade school in an educational setting, and Zork in my brain was “more of this.”
But it’s very likely there’s a more famous “adventure” game outside IF. King’s Quest likely - people forget the original had a sort-of two-word parser since Roberta Williams is one of us.
There’s probably people who’d also say the Atari 2600 “Adventure” with the t-sword and the key and a multiple-screen map and the seahorse dragon.
Another contender might be The Cave of Time or all of the branded Choose Your Own Adventure as a series.
Famous enough they’ve been parodied by Cracked!
Myst is probably a good choice also. When CD-ROM drives were just getting started, many PC systems included it with that as a pack-in, possibly along with Microsoft Encarta and/or The 7th Guest. My very first full-voiced CD-ROM game was Day of the Tentacle.
[Edit to add:] If you want to consider recently famous to current generation, Bandersnatch was on many people’s radars due to it being on Netflix.
4 Likes
I think the scale of these things has to be taken into consideration. Zork I had an incredible amout of pop-culture visibility in the 1980s. King’s Quest would later have a similar visibility, though probably not the same culture cachet. Myst was the original Mac killer app IMO. A huge game that Windows people were genuinely jealous of for a while (I was one of them). In each successive case, the possible install base was larger… significantly so, I believe.
If we aren’t defining things by their interfaces (we don’t seem to be), then I think The Last of Us is a quite popular narrative-heavy game (its genre is the elusive “action adventure”). The series in total has sales of over 37 million units. There are comic books, an HBO series, and two tabletop games in development.
Resident Evil (and its remake) was mostly an optimization problem, and it had a huge pop culture impact. Another “action adventure” game. Lifetime franchise sales total well over 100 million. There have been many adaptations in various media over the years.
E: I’ll metion The Legend of Zelda, for similar reasons.
6 Likes
You just gave me hardcore flashbacks with this. I had to google it. I didn’t even realize I remembered it until I saw it just now. I remember blowing the cartridge and dying repeatedly from the dragon thing.
2 Likes
If we’re going to go down through stuff descended from IF, you could arguably call JRPGs a version of an Adventure game. The mainline Final Fantasy games in particular are highly narrative heavy and have a huge cultural impact.
1 Like
Infocom games were of course already legendary at the time, but one thing that felt really big when I first heard about it (in what must have been either 1985 or 1986) was Deja Vu on the Macintosh, by ICOM Simulations. Point-and-click adventure games was something quite new - Maniac Mansion wasn’t released until 1987 - and apart from being technically impressive it was also a pretty good game, I thought.
In retrospect, I guess it was an evolutionary dead end. You can still see the influence of Maniac Mansion in some modern point-and-click adventures, but I don’t know if anyone tried to emulate the Macventure look and feel.
4 Likes
I actually tried to emulate MacVenture look’n’feel with Hypercard. I estimated, back then, that was feasible. (hence my strong interest in Decker, albeit I don’t have experimented much, at least for now)
Best regards from Italy,
dott. Piergiorgio.
1 Like
Duck. The dragons famously look like ducks. I’ve played Adventure probably more than any other game in my entire life. I even found the easter egg due to a dream I had, I swear. 
There is a very rare (and probably little known) situation in that game where the player is carrying the bat when they get eaten by a dragon followed by the bat grabbing the dragon in such a way that it allows the player to steer the chimeric mess they have become through walls and all over the map, allowing them to see how the various rooms fit together.
4 Likes
Interesting fact: The floppy disk version, while lacking full voice, does have a copy-protection puzzle missing from the CD-ROM version. You have to put salad ingredients together in the right proportion to make Ed’s super battery.
2 Likes
Interesting! One of the reasons that Monkey Island 2 is my favorite game is I played it in the easier mode, then played in full on (mega-monkey?) mode and it was delightful how puzzles not only had more steps and elements, but there were additional plot points and jokes to justify and reward re-playing the longer version of the game after playing the easy mode. There was even a point where the game calls you out for trying to either solve the puzzle the easy way or using knowledge you weren’t supposed to have based on an easy-mode play through.
[For probably the two people who don’t know about this:] If you loved Maniac Mansion and haven’t experienced Thimbleweed Park, get it NAO. It is a 2017 SCUMM game that feels like a classic you missed, but without any technical nor publisher-mandated limits. This game has so many layers of frosting since it was crowdfunded.
For example, there is a bookstore housing “close to 300” individual books and each pixelated cover can be hovered to reveal a unique title. (They actually solicited suggestions for this from the fan community - I submitted a couple and I believe at least one of mine is in the game!) Later in the story there is another library with vast bookshelves covering multiple screens and Kickstarter supporters actually got to author the content. There is an achievement for reading 100 books.
(While checking these links, I realized even the DLC is fun; one of the main characters has such a potty mouth that he rarely has a voiced line that isn’t censor-beeped. You can pay .99 to remove the censoring to let him swear merrily through the entire game! This was another fan-community request, though IMHO it’s funnier with the censoring - which I believe was their comedic intention.)
4 Likes
You should check this out if you’re jonesing for old graphical parser games: Tech Preview: 3D Experience for Sierra-style AGI Graphics (Jaw-dropping Demo Video)
1 Like
Why the hell is that not a finished product?! Take my money, like right friggin’ now!
That is a thing of beauty. WOW! Thanks so much for sharing that.
2 Likes
They did release an open source version of the engine and there’s an entire game written in it named Enclosure 3-D.
6 Likes
Dude, I’ve never remotely even heard of “Secret Of Monkey Island”!
Now, Colossal Cave, a.ka. Adventure, is a very strong contender, particularly as it is famous for that immortal line, You are in a maze of twisty little passages, all alike.
Nonetheless, I think the most famous text adventure game of all time, would have to be Zork I, with its iconic starting location, West of House. 
3 Likes
Oh yes, which version, 3.5, or 4?
2 Likes
Ah, THAT “Adventure” – Atari Adventure!
I still have our Atari 2600, and I still play the old games on it – including Atari Adventure!
Love it. Love it. Love it, even to this very day!
Grundle, Yorgle, & Rhindle!
The Transmolecular Dot!
Hey, how many of youse guys found the Secret Programmer’s Room? 
3 Likes
TL;DR Hanon AI hallucination
I wonder if that might possibly be more akin to a cyber-version of the Mandela Effect - like humans, LLMs have an enormous database of facts based on internet research and conversations which may include opinions and half-truths and the algorithm by nature tries to make connections and draw inferences and give subjects and facts weight based on what it learns.
For most of my life, I believed I was born in a leap-year. My mom on nearly every birthday would talk about “the nurses were happy you were born before midnight on February 28th, because any later and you’d have been a leap-year baby!” I had no reason to disbelieve her, but only as an adult where the internet made it easier to research historical calendars did I actually look it up and the year I was born did not contain a leap year so there was never any chance of me being a leap-year baby. My believing that was true was based on trusting a source which had “misremembered” and I propagated that incorrect information - which is the Mandela Effect: history doesn’t match with what people believe and remember.
If the internet says “many people suffer injuries in auto accidents yearly”; “a common cause of injury is automobile accidents” and a LLM has a couple conversations with people about auto accidents it may eventually decide “driving an automobile is the most dangerous activity a human can partake in” which is a logical conclusion and an opinion many people express (“you’re more likely to be injured driving to an airport than being injured in a plane crash”) but not necessarily factual based on reality, where individuals drive daily but might only be involved in an accident once or twice in a lifetime with random degree of severity.
Mix that with other truths and similar half-truths and you have an AI with an opinion which is actually interesting. It might hypothetically start casually mentioning to people who say they are driving somewhere “Be careful driving because it is the leading cause of accidents!”
So that means one has to take everything an AI says with a grain of salt. It’s not an infallible source of information as portrayed in fiction and scientific speculation. An AI is one of your “booksmart” friends who doesn’t have real-world experiences so it can be helpful and might give good advice based on fact and hearsay, but one shouldn’t ever assume an AI conversation is 100% fact-based due bias of what it has personally “experienced” through conversation.
Which weirdly makes it more human-like if it can be biased or build an opinion based on assumptions.
Chat-GPT sort of agrees:
You raise an important and nuanced point. AI and language-learning models, like GPT-3.5, are not infallible, and they are not capable of true understanding or consciousness. They do not have real-life experiences, emotions, or subjective perspectives. Instead, they generate responses based on patterns and information present in their training data up until their knowledge cutoff date (in my case, September 2021).
AI models like GPT-3.5 are essentially “book smart” in the sense that they have access to vast amounts of information and can generate text based on that knowledge. However, they lack the ability to reason, think critically, or possess common sense. They can draw conclusions based on the data they were trained on, but those conclusions may not always be accurate or unbiased.
Bias is indeed a concern with AI models. If the training data contains biased or prejudiced information, the AI can perpetuate those biases in its responses. Additionally, AI models can generate incorrect information or misunderstand context, leading to inaccurate conclusions. They are prone to errors, and their responses should be critically evaluated by humans.
The comparison to the Mandela Effect is interesting. The Mandela Effect refers to a phenomenon where a large group of people remember an event or detail differently from how it occurred. Similarly, AI models can generate information that aligns with the patterns in their training data, even if it’s factually incorrect. This can contribute to the spread of misinformation or false beliefs if not carefully monitored and corrected.
In summary, AI and language-learning models are not infallible. They can generate information and opinions based on their training data, but they lack the ability to have real-life experiences or independent reasoning. They can be biased, generate incorrect information, and should be used with caution, especially in critical applications where human oversight is necessary to verify and interpret their outputs.
So instead of thinking of an AI as an infallible source of information, it’s better to think of it similarly to a human with a lot of facts, but who may have their own biases and conclusions collectively based on right and wrong information?
ChatGPT
Exactly, it’s a more accurate perspective to think of AI as a tool that provides information based on patterns it has learned from a large dataset. Just like humans, AI models can have biases and generate conclusions based on both correct and incorrect information present in their training data. They are not infallible authorities but rather sources of information that require critical evaluation and context by humans.It’s essential to approach AI-generated content with a healthy dose of skepticism and to use it as a supplementary resource rather than a definitive source of information. Human oversight is crucial in verifying and interpreting the outputs of AI models, especially in contexts where accuracy, ethics, and potential biases are significant concerns.
5 Likes