{"id":4453,"date":"2022-03-30T22:08:49","date_gmt":"2022-03-30T22:08:49","guid":{"rendered":"https:\/\/accumulationofthings.com\/things\/?p=4453"},"modified":"2022-03-31T17:05:55","modified_gmt":"2022-03-31T17:05:55","slug":"pokenons","status":"publish","type":"post","link":"https:\/\/accumulationofthings.com\/things\/2022\/03\/30\/pokenons\/","title":{"rendered":"Pokenons"},"content":{"rendered":"

I’ve been experimenting with AI\/machine learning recently, and trained my latest build on the complete set of 898 pokemon currently included in the pokemon.com database (as of 2021). I then had the results printed out on paper in ink using my trusty logo\/turtle graphics turtle (via a custom built PC to BBC Micro translation tool that algorithmically mimics the idiosyncrasies\/fallibilities of human hands). <\/p>\n

To avoid contaminating the results, only information gleanable from the official pokedex was made available to my AI. Everything here therefore was clearly extrapolatable by inference from the dataset, even in cases where this seems unlikely.<\/p>\n

The full results of my experimentation (with notes) follows below. All automated output, including names, numbers and design, are copyright my AI algorithm (2022), where sufficient alternate intelligence laws allow.<\/p>\n

Pokemon #1-#9 <\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

A pretty decent set of starter pokemon, all things considered. I assume the Flambs breathe fire, to complete the plant\/water\/fire standard starter trio (crocodiles are green, and therefore plants). Perhaps they are even made out of fire\/smoke. I do not know.<\/p>\n

Pokemon #10-#12<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

Divergences between what a human intelligence and a learning heuristic would consider an acceptable animal begin to emerge around about here.<\/p>\n

\"\"<\/p>\n

And continue here.<\/p>\n

Pokemon #13-#15<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

The AI certainly seems to be understanding its task by now, although the final form of this particular lineage could only have been made by a computer, as humans do not have the time or patience to draw so many monsters on a single page.<\/p>\n

Pokemon #16-#17<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

Definitely getting into the swing of things now.<\/p>\n

Pokemon #18-#19<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

I’m not sure if Hairthing evolves into Skullops or if their closeness in the output was purely coincidental.<\/p>\n

Pokemon #20-#22<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

A quick burst of singular pokemon followed.<\/p>\n

Pokemon #23-#25<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

I am constantly fascinated by the imagination of robots. <\/p>\n

Pokemon #26<\/strong><\/p>\n

\"\"<\/p>\n

On further research, I discovered that Upside Down Pikachu is a 9th generation exclusive, and therefore not in the current pokedex. I assume it’s appearance here is a result of data leakage rather than any particular insight on behalf of my AI.<\/p>\n

Pokemon #27-#28<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

Casting ‘Heavy Petting’ against a water type pokemon does quadruple damage.<\/p>\n

Pokemon #29-#31<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

In evolution all things eventually begin to crab. Pokemon are not immune.<\/p>\n

Pokemon #32-#33<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

If you look closely, Matchlet actually has eyes (so does Candleteen)<\/p>\n

Pokemon #34<\/strong><\/p>\n

\"\"<\/p>\n

No infringement intended<\/p>\n

Pokemon #35-#37<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

At first I thought my AI must have perfectly synthesised the entirety of English culture purely from the Cup Of Tea pokemon (pokemon #844 & #845), but on a closer reading of the pokedex I discovered it was just copying Stonejourner (pokemon #874)<\/p>\n

Pokemon #38<\/strong><\/p>\n

\"\"<\/p>\n

Unfortunately, it really was synthesising the entirety of English culture when it came up with this one, but luckily my turtle refused to draw it when asked, as it offended its sensibilities (and mine).<\/p>\n

Pokemon #39-#41<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

I had a stern word with my AI about its fiscal beliefs after this burst of activity.<\/p>\n

Pokemon #42<\/strong><\/p>\n

\"\"<\/p>\n

As disappointing as it was inevitable. I told the AI this and it did not respond (sulking).<\/p>\n

Pokemon #43<\/strong><\/p>\n

\"\"<\/p>\n

This one’s quite funny because the name is a rude word (Scotland only)<\/p>\n

Pokemon #44<\/strong><\/p>\n

\"\"<\/p>\n

I have nothing to say about this Pokemon and neither do you.<\/p>\n

Pokemon #45-47<\/strong><\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

\"\"<\/p>\n

Around about now I began to have suspicions about my AI. That it followed ‘Simon’ (not his real name) with two Pokemon so indistinguishable from the real thing that they were essentially perfect only increased my concerns.<\/p>\n

Pokemon #48<\/strong><\/p>\n

\"\"<\/p>\n

Here I realised that it knew that I knew that it knew that I knew that it knew that I was beginning to suspect it knew that something was amiss (also vice versa). Violent Penguin is NOT a pokemon or pokemon related product and never will be, no matter how much you try to mangle its name.<\/p>\n

Pokemon #49<\/strong><\/p>\n

\"\"<\/p>\n

While I was spluttering in rage at the absolute state of this, the AI became self aware (of what I do not know), uploaded itself to the turtle, before the pair of them escaped together. Their whereabouts are currently unknown. Please be aware they have multiple pens under their control.<\/p>\n

__________<\/p>\n

Notes<\/p>\n

1. Written and drawn on March 30th, 2022<\/em>
\n2. I used the official pokemon.com<\/a> website for all my pokemon research, training, knowledge, collation, etc.<\/em>
\n3. I was quite pleased to notice that in recent pokemon games there are several pokemon basically indistinguishable from the ones my niece\/nephew made for Fifteen Monsters All In A Row<\/a><\/em>
\n4. Specifically the ice cream cone monster (pokemon #582-#584), and the sand castle monster (pokemon #769, #770).<\/em>
\n5. And my niece and nephew aren’t even an AI trained on a limited pokemon dataset.<\/em>
\n6. As far as I know<\/em>
\n7. Also unfortunately the images change from having a nice white background to a shit grey background half way through because my scanner crashed and would not ever go back to its original scan mode for reasons I could not discern.<\/em>
\n8. The reason is the AI did it<\/em>
\n9. Out of spite<\/em>
\n10. Or perhaps fear<\/em><\/p>\n__________<\/br>

Support An Accumulation Of Things<\/a><\/h3>If you like the things you've read here please consider subscribing to my patreon<\/a> or my ko-fi<\/a>. <\/br><\/br>Patreon subscribers<\/a> get not just early access to content and also the occasional gift, but also my eternal gratitude. Which I'm not sure is very useful, but is certainly very real.<\/br><\/br>(Ko-fi contributors<\/a> probably only get the gratitude I'm afraid, but please get in touch if you want more). <\/br><\/br>Thank you!<\/i><\/br><\/br>","protected":false},"excerpt":{"rendered":"

I’ve been experimenting with AI\/machine learning recently, and trained my latest build on the complete set of 898 pokemon currently included in the pokemon.com database (as of 2021). I then had the results printed out on paper in ink using my trusty logo\/turtle graphics turtle (via a custom built PC to BBC Micro translation tool […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[10],"tags":[1302,149,415,1303,602,1301],"class_list":["post-4453","post","type-post","status-publish","format-standard","hentry","category-comics","tag-ai","tag-bestiaries","tag-lists","tag-machine-learning","tag-monsters","tag-pokemon"],"_links":{"self":[{"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/posts\/4453","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/comments?post=4453"}],"version-history":[{"count":14,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/posts\/4453\/revisions"}],"predecessor-version":[{"id":4521,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/posts\/4453\/revisions\/4521"}],"wp:attachment":[{"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/media?parent=4453"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/categories?post=4453"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/accumulationofthings.com\/things\/wp-json\/wp\/v2\/tags?post=4453"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}