Search results

  1. moriad

    Gemini 3.5 and Antigravity come to Google NotebookLM

    I've been experimenting with Notebook LM on a subject about which I am the world's number 1 expert: my own fiction writing. (I use it only to analyze my writing. I've yet to find an LLM that is up to the task of actually generating high enough quality artistic text, and certainly none that can...
  2. moriad

    Police boast of hacking VPN where criminals “believed themselves to be safe”

    Their argument that being identified/identifiable stops antisocial and criminal behaviour is also demonstrably false. If it were true, just putting up CCTV cameras everywhere would prevent crime. For as long as there have been human beings, people have been doing terrible things to one...
  3. moriad

    Men use “vocal fry” more than women, counter to stereotype

    Most obvious and irritating vocal fry I've ever heard is Sam Altman's voice.
  4. moriad

    To teach in the time of ChatGPT is to know pain

    Education has value, but apparently students aren't seeing it. And why is that? Because universities have made it clear that they are profit seeking entities designed to squeeze every last cent out of students and their families while education is treated as an accidental side-effect of the...
  5. moriad

    Nvidia CEO tries to explain why DLSS 5 isn’t just “AI slop”

    Well, YouTube commenters are one audience. And they've spoken. But when I look at Nexus Mods and see screens full of pimped up sex dolls running around in negligees and high heels, with anatomically impossible assets set to full jiggle, I'm not so sure that YouTubers are quite in step with the...
  6. moriad

    How often do AI chatbots lead users down a harmful path?

    Oh, certainly. Other people have always been the modern human's biggest enemy. But... There's a physical and emotional upfront cost involved with dealing with other humans that LLMs lack, which makes them particularly inviting and insidious; plus their engagement never ends. The self-help text...
  7. moriad

    How often do AI chatbots lead users down a harmful path?

    The primary safeguard would need to be the other people in your life. Unfortunately, I suspect, the main reason people are turning to AIs is that they lack (or believe they lack) other trustworthy people in their lives. So here we are: If we had bacon, we could have bacon and eggs. If only we...
  8. moriad

    How often do AI chatbots lead users down a harmful path?

    I think you've cut to the heart of it. At least when it comes to the sycophantic stuff, I've learned that if an LLM ever gives me an even mildly negative response, I might want to take a much closer look.
  9. moriad

    How often do AI chatbots lead users down a harmful path?

    I've integrated Gemini Pro into my workspace stuff. It's extremely good at summarizing and drawing from multiple documents. (Though, of course, and crucially, it does make mistakes with matters of fact or statements made within documents). But we're talking here about documents full of random...
  10. moriad

    Having that high-deductible health plan might kill you, literally

    No, we don't guffaw, because the way things are going politically, some countries have strong lobbies trying to persuade the voters that the American system is better than what they currently have. (To some degree in Canada, much more so in the UK). Or if not the American system exactly, then...
  11. moriad

    10 things I learned from burning myself out with AI coding agents

    I think exactly the same things could be said for using LLMs to write text. Left to their own devices, the output appears functional, but is, in reality, clunky, inelegant, repetitive, and derivative, and the longer it is, the more narrative-breaking "bugs" you'll find, until the entire thing...
  12. moriad

    Being too nice online is a dead giveaway for AI bots, study suggests

    Advertisers are hoping to use AI to inject subtle under-the-radar advertising in the form of social media comments, not unlike the ones we've both been reading. I mean, I'm not sure how you do that in this case. But a lot of social media posts ask for advice, and frequently the advice sought...
  13. moriad

    LLMs show a “highly unreliable” capacity to describe their own internal processes

    Hahaha. I've been in meetings like this. Came to the same conclusion as you. Edit: Sounds like what happens when you "engage" with the system prompt only (which is also not a bad analogy for how many of us treat meetings.) Jungle Book - Vultures
  14. moriad

    Real humans don’t stream Drake songs 23 hours a day, rapper suing Spotify says

    Not necessary. The article mentions that you can sign up for an ad-supported account without a credit card. So it sounds like there is a free tier that is full of ads... streamed by mostly bots.
  15. moriad

    “Unexpectedly, a deer briefly entered the family room”: Living with Gemini Home

    People are still buying car alarms that go off for hours because a pine cone fell on their car.
  16. moriad

    YouTube denies AI was involved with odd removals of tech tutorials

    I don't know if anyone else ever puts subtitles on when they're listening to a video. If those subtitles are auto generated, it's quite alarming how much they actually censor the script. For example, I was listening to a video about characterization and story structure in films. And in an...
  17. moriad

    Ars Live: CTA policy expert explains why tariff stacking is a nightmare

    This is the only way to weaken Trump’s power in the next 18 months. And that kind of thinking is why we're in this mess. The GOP has an iron grip on power right precisely because they don't think that way. They realized a very long time ago that state legislatures, governors, local and country...
  18. moriad

    Seven things we learned from OpenAI’s first study on ChatGPT usage

    My partner has read somewhere between 18k and 30k novels. His ability to take a totally trash plot and morph it into something I can get excited about is epic. I usually don't even realize that my original plot is hopelessly cliche until he points out, with examples, just how many times he's...
  19. moriad

    ChatGPT may soon require ID verification from adults, CEO says

    Which would be great right up until the cops do a "wellness check" and end up shooting the victim who may not even be suicidal at all (to wit: a novelist using ChatGPT to create a profile of a fictional depressed character). AI's summoning law enforcement to people's houses doesn't strike me as...
  20. moriad

    ChatGPT may soon require ID verification from adults, CEO says

    So the day before your birthday, you're given guardrails to try to ensure that an AI chatbot won't do its damedest to persuade you to kill yourself, and the next day you're completely on your own. As though leading a person to suicide is only a crime if that person is a minor. Waiting for the...