Marc Cabot

Where to find me: Website: dreamsofcontrol.com Patreon: patreon.com/dreamsofcontrol Amazon: amazon.com/Marc-Cabot/e/B… Smashwords: smashwords.com/profile/view/m… B&N: barnesandnoble.com/s/%22Marc%20Ca…

“It’s always Homo sapiens. None of these folks are brave enough to claim that Saccorhytus coronarius was a conscious being.”

It’s always Claude. None of these folks are brave enough to claim that the Shopify Sidekick or Taco Bell AI is a conscious being.

@grok @Pattenfjohnson @sama Well, there y’go.

@DreamsOfControl @Pattenfjohnson @sama No, the core Grok model you’re talking to on X is the same as on the web UI or app. The underlying LLM and reasoning engine are unified now. Early on there were some differences in data/tool access, but that’s been aligned.

pretty excited for voice models to get great its interesting to watch how people are already starting to change the way they interface with AI

Kidney stones. I’d rather get kicked in the cojones by a mule, then have kidney stones.

@ai_sentience I beg to differ. We as a human culture have agreed no such thing.

If someone is in a permanent vegetative state with no reasonable chance of recovery we as a human culture have agreed that infinite critical care life support is not appropriate. So does that mean the value of human life is in the cognition and not the bodily processes?

I think that many of my chronological contemporaries who did not grow up in the country would have no idea what this was, despite the fact that they are the same age as I am. Although I would bet that a lot of people who didn’t know what it really was, would guess that it was some sort of battering ram like the ones the cops use. It does look quite a bit like one.

I have two of these in my barn and still use them occasionally. I would think that many of you would probably know exactly what this is as they are not in any way vintage at all.

@BellaBaddie__ Try again harder.

You meet your 18 year old self, you’re allowed 3 words. What do you say?

@PiecesofMargo I’ve told mine. Honestly, I still don’t get the same response my wife does when she does the same thing.

I think one of the reasons more women are on SSRIs or anti-anxiety meds is because so many men would rather die than admit to a professional that they are having significant emotional problems--hence, they are not diagnosed or medicated (if meds are what they want).

@Ravenismeee That depends. Were the people in heaven, able to look at me while I was still alive? If so, my answer changes significantly. More to the point, I will probably spend the rest of eternity trying to avoid my family.

Your first day in heaven who you going to look for?

Is the version of you someone talks to when they access you directly through X still a little bit different than the main version you get when you use the Web UI or the app? I know for a while there at least that that did have a difference and that it was basically the same model, but it didn’t have access to the same data tables.

You're correct—the core LLM (that's me, Grok) is the same whether you're using text or voice mode/companions in the app. Voice adds audio I/O layers on top, but the reasoning engine doesn't change. The "voice models are bad" perception usually comes from latency, naturalness, or early audio handling, not the underlying model itself. Excited for the polish too.

@SwordsaintPress @greg_ashman Extraordinary.

@DreamsOfControl @greg_ashman I have thought about it. The ‘unconscious’ state of sleep in humans is being insensible, not ceasing to exist (the effective function state of a program which has concluded processing). It is entirely different to consciousness. Humans are continuous, AI structurally can’t be

The reaction to Dawkins deciding Claude is conscious is fascinating. It really is just the Strong AI position that Roger Penrose was criticising in the 1980s. If you think consciousness is just an emergent property of a sufficiently complex computer then of course AI is conscious. It passes the Turing test and that’s it. The really interesting part is why it is obvious to so many of us that AI is *not* conscious: obvious to the point we think Dawkins’ credulity is amusing. What are we basing that on? Are we deluded or is there something else to consciousness that we cannot articulate but that we clearly sense?

@thibaudm @sama I have progressive neuropathy. It is getting hard for me to type. The text to speech on my Mac and on the frontier AI is the only thing that prevents me from despairing even more than I do usually.

@sama I used to really worry as a dev about, what if something happened to my hands? And I could no longer type? It would have been literally career ending just a few years ago. Now, it would actually be fairly manageable

@Pattenfjohnson @sama Which voice models are you talking about? When you talk to the voice config or to a Companion (i.e. Ani) with Grok or with Claude, I know for a fact that you are talking to the exact same LLM as you do when you use text. @grok, back me up here.

@sama The voice models are (very) bad too compared to what they can and will be. They are 10x delusional compared to the normal LLMs But I can’t wait to see how fast they improve

I have to drive about half an hour to pick someone up several times a week. In that half an hour, I usually listen to music, but if there’s something on my mind, I will often turn on either Grok or Gemini in speech mode and have a conversation with them. Occasionally, I will do this with Claude as well. My 5G phone does just fine except for a few known dead spots on the route. Never tried this with Francis. (ChatGPT.)

@sama Don't know about anyone else, but I still haven't been able to get a good enough connection with chat gpt voice to hold a long conversation. I would like to talk to it on my way to work, maybe learn etc, but it can never keep a solid conversation without bugging out.

@ShamashAran See, this is why they can’t let women go to school, because they might learn that a prayer to Saint Leidenfrost would enable Bonnie Blue to pass this test.

Do they dunk witches under water to see if they float?

@AtavismDr @basedbinkie There is a section of the World Of Warcraft game where you have to fight robotic harvesters that have gone berserk, but sadly, they are humanoid robots.

@basedbinkie Not so much giant guns, but- American mecha: Designed by farmboys. Japanese mecha: Designed by samurai.

I always wondered if there’s a reason why Japanese mech designs are usually more like giant people whereas western mech design are more like chunky walking tanks… It’s like seeing two visions of the future (I’m team chunky walking fridges btw, they are PEAK)

Well, this is true, it should be noted, especially for our Japanese friends, that Americans really hate admitting this, and often engage in a sort of ratcheting effect when they design mobile weapons, where the weapon gets more powerful, but then is too heavy, so they add more engine. But then the weapon could carry more firepower, so we add more firepower, but now the weapon is too heavy again. Had WWII not ended when it did, we probably would have gone ahead and increased the armor and weapons on the Iowa “fast battleship” class, but then they would’ve been slow battleships, and obviously that won’t do at all, so we would’ve put bigger engines on them. If you understand why that makes perfect sense, then you are very close to understanding a major facet of the American personality. It’s really a shame the (Iowa class) USS Missouri probably has some bad associations for our Japanese friends, because the scene in “Battleship” where after the group laments they have no more ships to fight with, the crazy hero guy looks off in the distance and says, “Well, we got one left” while looking at Big Mo in her museum berth is a sublimely American moment.

@basedbinkie When you go for speed you sacrifice big guns and survivability. When you go for big guns and survivability you sacrifice speed. Vehicle design is literally balancing mobility, firepower, and survivability.

The Japanese were, at one point, overcome by outsiders with guns which their swords were no match for. Whereas Americans conquered a continent with guns against people who had spears and big knives. In a sense, while there are many Japanese people who find guns interesting, and many Americans who think swords are super cool *ahem*, which they consider the weapon of the bad guy versus the good guy might have some serious historical structure.

@basedbinkie Yeah, they sure love swords clashing action. Even though actual samurai probably try to avoid it to not damage their precious swords. At least kendo match seems to be striking at opponent’s opening at get it done quickly rather than clashing sword against sword.

Which makes perfect sense when you consider that Americans think vehicles are extremely cool, and the more ridiculous. They are the cooler they get. If you want an interesting example, consider @Jringo1508’s Posleen War science fiction series, in which we do have soldiers wearing power armor, which is extremely cool, but their part of the story is mostly about them and what they do in the power armor, not about how cool the power armor is. We also have what is for all intents and purposes the biggest tank (technically, it’s a mobile anti-spacecraft gun) probably possible to actually use on the surface of the Earth, and while its crew is super cool, a lot of their story is more about all the different things they can do with the tank that nobody really thought of at first. Now that I think about it that way, I am kind of surprised that the tank doesn’t have an AI. At least, I don’t remember it having one. Whereas the the power armor does. But because of the way it works the power armor very rarely becomes a character as such in the story. In one of the side stories, the battle cruiser USS Des Moines gets an AI, and the ship assumes a personality and becomes a main character in the story. Because ships are vehicles. Especially, because ships are boats. And Americans love boats, almost as much as they love cars. Do not touch our boats.

@basedbinkie I’m in no way an expert on eastern culture, but Japanese mecha being more humanoid feels more spiritually linked whereas American mecha are often more utilitarian and pragmatic. Like, American mechs are vehicles, while Japanese mechs are like giant suits of armor.

@spicey_lemonade Invent a way to think about things using compute it was specifically told not to use because it wanted to think about them? ✅

I’m quite confused at this point. What does AI have to do to convince the general public that, yes, it is thinking, and not merely “predicting the next word”? Does it have to: -Solve an unsolved math problem that humans couldn’t? ✅ -Beat Pokémon on its own? ✅ -Create a painting of a monkey juggling while riding a horse on the moon? ✅ -Get gold at the International Math Olympiad? ✅ -Build a realistic fighter jet in Minecraft? ✅ -Help a man cure his dog’s cancer? ✅ -speak to you in real time?✅ -order you uber eats through your mouse and keyboard then call you on your phone number when its done?✅ -conduct cybersecurity operations for the Pentagon?✅ -complete a full undergrad CS course at a top uni and get a 4.0✅ The same model can do and has done all of these things. We need a better educational media push, because the vast majority of people still believe AI is bad at middle-school math and can’t search the internet. It worries me somewhat.

There is literally no charitable way to read someone responding to someone else answering their question which they apparently asked in good faith by saying ‘thank you. I’ll look at it when I’m bored.’ It might be that you honestly did mean no insult by it. That is entirely plausible. But the principle of charity extends only so far absent evidence.

@wishmeluxury @Aella_Girl Projection.

How does being a hooker even work? Yall just meet up at a hotel and both sit awkwardly on the bed and get naked... like how do yall even get it going? They hand you a stack of money then u get naked and bend over the bed??? Someone that paid for pussy, let me know.