What Chat GPT thinks of the Standard

27 comments on “What Chat GPT thinks of the Standard ”

• Comments are now closed

1. Muttonbird 1

   18 February 2023 at 10:49 am

   Not left wing, just accurate.

   Also scarily good. While the essay on The Standard does repeat some of the same ideas, it is informative and easy to read.

   I can see why educators are so concerned. Soon we will not have the people capable of inventing AI because AI has undermined learning…

   Skynet, here we come.
   • lprent 1.1

     19 February 2023 at 1:08 pm

     Does way too much emphasis on me. I didn't 'found' The Standard. I got drafted into supporting the site technically. Still chained to that role.

     I didn't write posts for quite a while after startup, and over the last 15 years, 6 months, 3 days since the site started, I have only managed to publish 962 posts.

     Roughly one every 6 days (and a lot less recently). Although I see I have now done 23,812 of the 1,823,077 approved comments on the site.

     Mostly I write lot of code rather than the lots of political posts and comments. My demanding critics are the build pipelines, colleagues in whatever agile team I'm in, and the customers of my employers.

     Whereas Mickey has published 2450 posts in 8 years. Nearly as many as Anthony R0bins, who published 1889 over 6 years – and 605 as r0b over 2 years earlier! Eddie published 1434 in 6 years.

     There are a lot of authors who wrote about half my posts in just a few years. weka with 436. advantage with 439, all your base with 687, the inflammatory IrishBill with 535.

     There are 89 authors who have published posts on the site. Some only a few, and as you can see some wrote quite a lot.

     By really the site runs mostly on the commenters and the moderators who spend endless hours making sure that the comments section doesn't go toxic. Without them the site isn't alive and most of the authors wouldn't bother to write without clear feedback.
     • Muttonbird 1.1.1

       19 February 2023 at 7:37 pm

       Good stuff, LP. We commenters are truely grateful for the initial and continuing work you do.

       But, despite your humility our new overlord, the AI chatbot, has apparently gathered all it needs to know and has placed yourself at the centre of proceedings.

       Would loved to have seen our new overlord, the AI chatbot, mention @mickeysavage because his tireless contribution to this forum keeps it relevant, ongoing, and away from the rabbit hole.

       Kia kaha.
2. weka 2

   18 February 2023 at 10:59 am

   In the early years of the blog, Lynn Prentice wrote most of the content himself, covering a wide range of political topics. However, as the blog grew in popularity, it began to attract contributions from other left-leaning writers and activists.

   Is that true? I was under the impression that there were a range of authors in the early years.
   • pat 2.1

     18 February 2023 at 11:42 am

     2007 archives suggest not…unless Lynn wrote under multiple pseudonyms
     • lprent 2.1.1

       19 February 2023 at 1:10 pm

       I have only used two. lprent and AncientGeek. The latter was mostly been used for testing, and I have written some comments with it. Mostly back in 2007/8. I was using lprent for moderating then.
       • Thinker 2.1.1.1

         20 February 2023 at 9:25 pm

         Beware of Geeks bearing GIFs…
         • lprent 2.1.1.1.1

           21 February 2023 at 9:01 am

           Sorry no Trojans here. I believe you can find some in our local space.

           https://en.wikipedia.org/wiki/Earth_trojan
           
           • Mike Smith 2.1.1.1.1.1

             21 February 2023 at 3:53 pm

             As one of the founders I can say that the fact that the Standard is still going and relevant is pretty well entirely due in various ways to Lynn and Mickey Savage, as well as the moderators and commenters. Huge thanks are due to all.
3. Lanthanide 3

   18 February 2023 at 11:02 am

   Is ChatGPT left wing?

   No. Think of it like a glorified search engine. It has no understanding of what it is writing, and likely it has simply taken existing articles that people have written about these blogs on the internet and synthesized them into a new document with excellent prose and grammar.

   See the timestamp in this video, by a computer scientist on the topic of ChatGPT writing malware: https://youtu.be/L9w_k2ypRr0?t=360

   And: https://www.theverge.com/23604075/ai-chatbots-bing-chatgpt-intelligent-sentient-mirror-test
   • AB 3.1

     18 February 2023 at 11:32 am

     Yes. It appears to be good at a fairly conventional and in the end rather trite synthesis of publicly available sources. As you say, it produces unobjectionable prose and grammar. Inevitably, the content is also quite dull and uninteresting. That is because the processes by which it works make originality impossible – it has no 'mind'. Best seen as as a very good plagiarism tool. It will be popular due to us deciding that education is solely a private good and university students should pay massive fees – placing enormous pressure on them to get a return on that 'investment'. In the end, we can blame the neolibs for many forms of depravity.
     • pat 3.1.1

       18 February 2023 at 11:50 am

       In the case of say journalists will anyone other than they themselves care however?
       • AB 3.1.1.1

         18 February 2023 at 2:35 pm

         I'd guess not. In some ways it's actually useful – a sophisticated search engine that gives you a potted summary of its hits, rather than just a list of things to waste your time on and make you lose the will to live. The mistake will always be to trust it.
         • pat 3.1.1.1.1

           18 February 2023 at 3:59 pm

           I imagine those looking for cost savings in the prose industries will be quite willing to trust it….and in most instances the audience will be unbothered.

           Perhaps we should be bothered, but history suggests we largely will not.
           • Macro 3.1.1.1.1.1

             18 February 2023 at 4:44 pm

             Interesting experiment in what you moot right here:

             • Rabbi Joshua Franklin told ChatGPT to write him a 1,000-word sermon about intimacy and vulnerability.
             • Franklin told his congregation he was "deathly afraid" when applause broke out after his sermon.
             • Franklin told the AP that ChatGPT is "really great at sounding intelligent," but lacks empathy.

             Franklin told the AP that his congregation guessed the sermon was written by really "wise, smart, thoughtful individuals." But he added that while ChatGPT may be intelligent, it cannot be compassionate and it cannot love — and so, cannot create the "things that bring us together."
             • AB …

               18 February 2023 at 7:48 pm

               Interesting. The question that comes to mind is – how could the rabbi detect its deficiencies and his audience could not? Was it simply that he knew it had been produced by ChatGP and set out to look for issues – while his audience though it had been produced by a (presumably trusted) rabbi and set out to look for spiritual nourishment? Or is it something more objective – that the rabbi just knows a whole lot more than his audience about the subject and what has been written on it?

               Of course, "wise, smart, thoughtful individuals" have written stuff on this topic. What is interesting is that ChatGPT's effort must have given more weight to these serious sources than to the mountain of dross that is no doubt also available. Some sophisticated weighting of this sort has to be a core component of its rule-base. All we really need is for Elon to buy it and twist its rule-base to favour right-wing, libertarian sources over all others.
4. Incognito 4

   18 February 2023 at 12:02 pm

   The TS URL is incorrect, twice.

   The verb “things” in the title is funny and clearly not written by a chatbot.
   • mickysavage 4.1

     18 February 2023 at 12:30 pm

     Yep the heading is my mistake, now corrected. The URL is Chatbot’s interesting that it should do this.
     • roblogic 4.1.1

       18 February 2023 at 1:18 pm

       ChatGPT is good at constructing plausible sounding language but it does not "think" symbolically or really have a category of "fact".
     • Lanthanide 4.1.2

       18 February 2023 at 3:41 pm

       It should be retitled: What ChatGPT Has "Learned" The Internet Says About The Standard.
5. Ed1 5

   18 February 2023 at 10:17 pm

   I have thought that The Standard benefits from a fair but fairly sensitive moderation policy; in contrast to say Kiwiblog which appears to operate a moderation policy of "if it is not illegal then it is OK" Rather than comparing with WhaleOil I would be more interested in the ChatPPT on thebfd which appears to be the successor to WhaleOil. Is calling Kiwiblog and thebfd the Sewer and the Cesspit unfair?
6. Ad 6

   19 February 2023 at 9:14 am

   I'm looking forward to ChatBot and other AI systems putting an ever-greater premium on human writers generating more concise, creative, compelling prose analysis of politics.

   AI writing is a washing machine or Electrolux mower. We've needed it, especially in social media.

   Hopefully it kills off the political biography permanently.

   I only have fears for the boorish and mediocre, not for the likes of Chauncey Devega or Heather Digby Parton.
7. Muttonbird 7

   19 February 2023 at 10:46 am

   Ran it through GPTZero:

   Your text is likely to be written entirely by AI

   The nature of AI-generated content is changing constantly. While we build more robust models for GPTZero, we recommend that educators take these results as one of many pieces in a holistic assessment of student work.

   Average Perplexity Score: 22.524

   A document's perplexity is a measurement of the randomness of the text

   Burstiness Score: 8.370

   A document's burstiness is a measurement of the variation in perplexity

   Your sentence with the highest perplexity, "It has also been involved in promoting progressive candidates and supporting left-leaning political parties.", has a perplexity of: 36
8. gsays 8

   19 February 2023 at 7:22 pm

   FWIW, I found this very interesting;

   He appears knowledgeable and a good communicator.
9. Thinker 9

   20 February 2023 at 9:13 pm

   I KID YOU NOT…

   I thought I'd take a leaf from Mickey Savage's book and asked ChatGPT who Christopher Luxon is. This is an unedited response. So, to all those who say Chat GPT isn't intelligent, I suggest you check out the prediction it is making…

   who is christopher luxon?

   Christopher Luxon is a business executive and former politician from New Zealand. He is perhaps best known for his role as the former CEO of Air New Zealand, one of the country's largest and most prominent companies. Luxon joined Air New Zealand in 2011 and served as CEO from 2013 to 2019, during which time he oversaw a period of significant growth and expansion for the airline.

   Prior to his time at Air New Zealand, Luxon worked in a variety of senior management roles at Unilever, a multinational consumer goods company. In 2020, Luxon announced that he would be running for a seat in the New Zealand Parliament as a member of the National Party, and he was subsequently elected to represent the Botany electorate in the 2020 general election. However, he resigned from politics in 2021 and returned to the private sector, where he is widely expected to continue playing a prominent role in New Zealand's business community.
10. Mike the Lefty 10

    21 February 2023 at 5:58 pm

    One of the best things about The Standard is that nutters and trolls seldom get on here to vent their delusions on everyone else. The standard of debate is usually quite high – in contrast to The Daily Blog where delusional hysteria rules the roost.
    • Incognito 10.1

      21 February 2023 at 6:29 pm

      I think we all harbour a delusion or two, deep down inside, as it is somewhat of a spectrum, IMO – call me delusional. Most of us tend to keep quiet about those and some are not even aware they have them. The problem is not so much the delusion per se but how well (as in: successfully) or poorly one can argue for it here and in good faith [no pun].

      Chatbots, on the other hand, don’t have illusions or delusions. They lack a contextual framework, understanding, purpose, and intention. They are highly sophisticated auto-complete algorithms without fact-checking, and literally mindless brute-force number crunchers that predict the most likely next word in the sentence/sequence.

• Comments are now closed

• Log in
• Donate
• Contribute post

• Articles RSS