The creators of ChatGPT have a far more lucrative goal in mind than simply taking my job. The bigger picture is the multi-billion dollar sector that is internet search. And that is why it has been dubbed the Google killer. Google’s parent company Alphabet made $104bn (£86bn) in revenue in 2020 alone, just from search. Taking even a tiny fraction of that market would be a huge prize… Imagine going to a search engine, typing in your query and back comes one definitive answer, rather than pages and pages of links (and ads) to wade through?
https://www.bbc.com/news/technology
Erm. No, I don’t want “one definitive answer”. The whole attraction of Google is that there’s a lot of different sources of information, not just one “final truth” preselected by some woke 23-year-old Snapchat addict. Also, Google has no barrier to entry while the chatbot is completely inaccessible unless you provide tons of info it can’t possibly need. Given how woke its creators are, the access to the chatbot and the quality of its responses will unavoidably be regulated according to everybody’s ideological reliability profile. People who are asking the chatbot facetious questions about political issues are putting themselves on all sorts of blacklists as we speak.
Let’s hope that the general stupidization of all online options doesn’t rob us of search engines, giving us instead “one definitive answer” chosen by a bunch of eager brainwashed children.
Clarissa's Blog 
Yeah, already sussed out this problem years ago: when I have an annoying health issue that my doc knows nothing about, I don’t want one definitive answer. That’s already available from the four or five drug-company-sponsored mainstream answer-your-medical-questions websites. And the information is completely useless unless you’re a hypochondriac trying to diagnose yourself with cancer, or you’re in the “just give me a drug for it” camp.
What’s actually useful is websites run by obsessive people who have the same health problem as me (eg Jenny Ruhl), who have helpfully compiled all the available research on it, in one place, often with handy plain-English summaries. These people aren’t experts in the traditional sense, and they get deep-sixed by the major search engines. They are also the people I’m most interested in reading, and over the past few years they’ve become more and more difficult to locate, which pisses me off because I know they’re still out there, it’s just that Google is snuffing them in favor of Livestrong.org and its ilk.
Bastards.
Getting an AI to do that… might as well just put up a huge brick wall between the public and any information we might ever find useful.
LikeLike
You are underestimating the laziness of current human race — it’s no different than the phenomenon of students cheating by using Internet resources though they have unprecedented access to online books and reading material (which the earlier generations never had, or at the very least involved trekking to the nearest library, going through the massive manual catalogues and then finding the desired text amongst tomes of related literature — good old days!).
The appeal of a chatbot which directly provides you refined search results and even puts them in a semi-coherent narrative must be tremendous. I always think it’s not the “growing artificial intelligence” capability but the “average human stupidity” that may make humans redundant and/or subjects of some disembodied machine.
LikeLiked by 1 person
God, I’m afraid you are right. Your comment inspired me to post a quote from Tolstoy that describes just this kind of people.
LikeLike
I have grave doubts even about the “semi-coherent narrative”. I find the chattybots’ output damn near illegible. I have no problem reading, say, comments here, no matter how long. It takes me twenty or thirty minutes and a herculean effort to get through a short paragraph generated by the infernal machines. It’s like reading insurance paperwork.
LikeLiked by 1 person
That’s exactly my feeling. Thank you for helping me realize I’m not crazy here. I literally can’t make my way past a single sentence. People go into raptures over how human these texts sound, and I simply don’t see it.
Writing in the sense of churning out sentences is easy. It’s finding people who would want to read your writing that’s very hard. After the initial sense of novelty evaporates, who’ll want to read this stuff?
Take this blog, for instance. Why are people coming here for years? Isn’t it because there’s a real person with a real life, real contradictions, real problems behind the text? And the readers are all real humans who share little bits of their real lives? If all this were generated by a machine, who’d want to waste time on it? Definitely not me, that’s for sure.
LikeLiked by 1 person
I’m curious why this is the case with the artificial writing, but I find it so repellent I can’t properly analyze what’s wrong with it. What is that? I’ve read typewriter manuals that were more engaging.
The rational part of me wants it to be a problem of logic and syntax: any device that’s simply looking at a bunch of examples and coming up with a mathematical average of them, isn’t holding any single thread of logic through a paragraph, or a page. It’s taking a whole bundle of actual people’s actual threads of logic, snipping them up and running them through a blender, and pouring the results out in text. There’s no continuity. If the second sentence elaborates on the first, that’s a chance meeting.
It’s like the way AI “art” handles hands (if you want to see something truly creepy, poke around for “AI generated sign language manual” images)– it doesn’t know anything about hands, it just has a bunch of pictures labeled “hands” and it’s trying to come up with some sort of visual average, that actually looks like a manual of congenital finger defects, or sign language for severe burn victims.
And yet. I’m not convinced it’s a completely rational issue. What is writing? What are we doing when we write? We’re communicating with other people. When we read, we are connecting, on some level, with the writer. Sometimes, I pick up a book, and it’s a slog getting through the first chapter, because the writer is unfamiliar, and not a style I’m well-versed in. Those initial bits are like the awkward getting-to-know-you phase of a relationship, where I get used to how someone talks, what their hangups and linguistic tics are. All of these things make them more human, and more known to you. Once I get used to them, the reading goes faster.
This is partly mechanical. But it’s also social and spiritual. It’s ink, pixels, words… but what happens is not physical. In hippie new-age language, you might say… at a bare minimum, the part of yourself that exists on the astral plane– where ideas and concepts exist– is reaching out and touching the part of the writer that exists there. Perhaps it is even possible to connect on a spiritual level, with some writers. However this process works, we do it all the time. When we read, we reach out to make a connection with the writer.
I think this is why insurance paperwork is so hard to read. There is no real author, only a set of protocols, combed over by lawyers, proofreaders, accountants… by the time it goes to print, so many people have made so many adjustments to it that most traces of humanity have been removed, and its fragmented voices, alternately forced and asphyxiated by the legal and technical requirements of the medium, leave little to connect to. I have read truly beautiful legal writing. But it invariably is written by one person, start to finish, an unbroken chain of logic animated by the will of the writer. Insurance docs though… (shudders). Too many authors, too many constraints (and unfortunately, brevity isn’t one of them). You reach out with your soul to find another soul to connect to, and get… nothing. People have been erased from it in all the cut-and-paste, like an old-fashioned ransom note clipped from newspaper letters. The headlines they were cut from are silenced, to communicate something inhuman, hostile.
AI can’t help but have the same problem, only larger. We are used to reading the thoughts of others. the AI has no thoughts. It is a horrifying conglomeration of snippets of thoughts, with no soul to animate it. We reach out in spirit to shake hands, find a yawning void (or perhaps a tentacled horror), and instinctively scramble back from the edge.
LikeLike
It’s a bit reminiscent of the clockwork doll story in Tales of Hoffman. My kids like the doll aria from the opera. I find the story weird and disorienting, and never really connected with it before, but… suddenly they’re a good metaphor and I like it.
LikeLike
“I find the story weird and disorienting”
If you’re talking about the opera, it’s great but was widely misunderstood for years. There were/are many revised versions and a terrible change in the order of acts, but around the 1980s they returned more to his original conception.
The opera is basically about male romantic attachments and how they change over time.
Olympia the doll is a man’s first love in youth, pretty but ultimately empty headed and incapable of growth or development.
Antonia is mature love as an adult with a sensitive woman who loves him until death does them part.
Giulietta the Venetian cortesan is the prototypical mid-life crisis girlfriend/mistress/second wife who’s vulgar and grasping, leaving him drained and defeated.
Stella is the idealized love he can never have (which is why she barely appears)
Nicholas/the Muse is his internalized feminine, instinctual side that enables him to create but ultimately keeps him isolated.
There’s more but that will get you started….
LikeLike
Nice. Thanks for the guide.
LikeLike
Dear Clarissa, like all intelligent people you take for granted that most people are intelligent. Newsflash: most people are NOT intelligent, at least not what you would consider intelligent.
LikeLiked by 1 person
How sad, though. It’s almost physically painful to imagine people for whom “one definitive answer” from a search engine is an attractive thing.
LikeLike
Thuh Beeb: “Imagine going to a search engine, typing in your query and back comes one definitive answer, rather than pages and pages of links (and ads) to wade through?”
Anyone else here a fan of the 1960s British television show “The Prisoner”?
It’s like whoever wrote that never had the pleasure of watching the episode called “The General” and the experience of watching all of those lovely Village denizens (including our ever-resistant Number 6) getting to enjoy being given all of their answers in a standardised format that differs between people not even in the slightest.
Jamie Zawinski of Netscape and Mozilla infamy (just look for “jwz blog”) had an amusing cultural reference: this is like that simulated parrot of Bryce Lynch’s in the Max Headroom series.
He’s absolutely right, this is all pseudo-intellectual grooming to be well-behaved synthetic parrots.
Avi: “… most people are NOT intelligent …”
That’s because most people employ an edge detection algorithm for any problem that allows them to consider it on its surface without considering it as a whole, and as long as the algorithm works, they have no need to apply anything even resembling intelligence directly on any subject.
For all of the problems that the MBTI poses, it’s incredibly useful as an edge detection algorithm for detecting people likely to employ edge detection algorithms.
And for your edification: that’s everyone who ticks the S box plus two E types that lean that way because of using extraversion as a proxy for sensing.
There’s an interesting population distribution pie chart that shows how frequently the six types that aren’t this way occur versus the ten that follow the edge detection pattern, and it’s interesting because it’s almost a Pareto distribution.
Most people aren’t intelligent, many of them aren’t even interesting, but one in five?
That’s actually much better odds than more cynically minded people would have expected.
LikeLike