Path: nntp.eternal-september.org!news.eternal-september.org!eternal-september.org!feeder3.eternal-september.org!usenet.network!news.neodome.net!rocksolid2!i2pn2.org!.POSTED!not-for-mail From: joes Newsgroups: comp.theory Subject: Re: The halting problem as defined is a category error --- Flibble is correct Date: Fri, 18 Jul 2025 21:42:03 -0000 (UTC) Organization: i2pn2 (i2pn.org) Message-ID: <357a5fa9acde7e986682a5071c2de5c254a2919b@i2pn2.org> References: <105bdps$1g61u$1@dont-email.me> <105c0lk$1k7ip$1@dont-email.me> <105c22v$1k9r9$3@dont-email.me> <105c5rt$1l4j7$1@dont-email.me> <105cddu$1r7mi$1@dont-email.me> <105e259$26kvp$1@dont-email.me> <105e8nt$288fm$1@dont-email.me> MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Injection-Date: Fri, 18 Jul 2025 21:42:03 -0000 (UTC) Injection-Info: i2pn2.org; logging-data="1205950"; mail-complaints-to="usenet@i2pn2.org"; posting-account="nS1KMHaUuWOnF/ukOJzx6Ssd8y16q9UPs1GZ+I3D0CM"; User-Agent: Pan/0.145 (Duplicitous mercenary valetism; d7e168a git.gnome.org/pan2) X-Spam-Checker-Version: SpamAssassin 4.0.0 Am Fri, 18 Jul 2025 14:53:31 -0500 schrieb olcott: > On 7/18/2025 1:01 PM, Mike Terry wrote: >> Well there you go - if you feed incorrect statements to a chatbot, it's >> no surprise it is capable of echoing them back to you.  Even Eliza >> could do as much... >> > The above definition of HHH is ALL that the bots ever see, and there is > no basis for anyone to determine that it is incorrect. HHH only detects that its self would get stuck in recursive simulation. >> You can't expect people to "acknowledge" false claims - I told you >> years ago that HHH does not detect any such non-halting pattern.  What >> it detects is your (unsound) so-called "Infinite Recursive Emulation" >> pattern.  I wonder what your chatbot would say if you told it: >> ---  So-called Termination Analyser HHH simulates its input for a few >> steps then decides to return 0, incorrectly indicating that its input >> never halts.  In a separate test, its input is demonstrated to halt in >> nnnnn steps.   [Replace nnnnn with actual number of steps] >> > I have proven that DDD simulated by HHH and directly executed DDD() are > in Claude.ai's own words are > "computationally distinct objects that have demonstrably different > behaviors." That's bad. Nobody is running all their programs through HHH, in which case they wouldn't even need to test them with HHH, since they are already running everything through it and it refuses to execute it. We are interested in DDD only. Nobody cares what HHH does to any program. > I tell you this: > "Halting is ONLY reaching a final halt state" > hundreds of times and you pretend that I never said it. Liar. Quote? >> How will you ensure CACM gives your paper to peer reviewers who are >> "not liars" [aka, reviewers who aren't concerned about correctness of >> your argument, and instead just mirror back whatever claims the paper >> makes] ? > No one even attempts yo point out any actual errors. > Joes just said that HHH cannot possibly emulate itself after I have > conclusively proved that it does. I didn't. You forgot the qualification. > I rewrote that today to make it easier to understand. > You are the only human in this group capable of actually understanding > what I said. Fucking speciesist. >> I doubt you'll have any luck tricking the reviewers at CACM.  Unlike > The most important reviewer at CACM did exchange 20 emails with me to > review my work. He ended up giving up because he did not know x86 > assembly language well enough. Oh really? Who was it? What did they say? When was that? > Even you made the ridiculously stupid statement that DDD correctly > simulated by HHH will eventually reach its own simulated "return" > instruction final halt state. Time and date please. -- Am Sat, 20 Jul 2024 12:35:31 +0000 schrieb WM in sci.math: It is not guaranteed that n+1 exists for every n.