Goatchaps

Naaa. Baaah. Meeeh.

[Goatchaps @Goatchaps]

If you are interested in day trading futures, Maple has an easy system. @swapmeet @CryptoConvoy

[MapleStax Trades @MapleStax]

Lucid Trading eval giveaway! ⚡️

How to enter:

-LIKE

-FOLLOW

-Tag 2 friends!

Make sure you are following me and
@TradingLucid, I will be checking!

Winner announced this SUNDAY!

[Goatchaps @Goatchaps]

@EricLDaugh Why don't they understand that "man" can mean someone from the race of man? 🤯

Instead, they try to argue that humans are female first in the womb, so everyone can be referred to as "her."

[Goatchaps @Goatchaps]

@MapleStax CBC

[Goatchaps @Goatchaps]

@MapleStax I'd like to take a look at it, please and thank you.

[Goatchaps @Goatchaps]

If you are using websites that supposedly check for AI content, AKA AI detectors, you have to read this. Those detectors are full of false positives, though there is one that seems accurate.

I get free stuff for writing Amazon reviews. Sometimes my reviews get rejected even though they follow the terms set forth by @amazon. I wondered why they would be rejected, and Amazon doesn't tell you why-- they simply reject it.

So I wondered if they thought I was using AI for my reviews. Let me be clear: I've never used AI for any of my reviews, not even to obtain information for me. I'm a sports editor, and I have been for about five years. I write my own content.

Anyways, I pasted some of my reviews that were rejected into AI detection tools to see if it misidentifies my work as AI. The results were amazing to me. The first rejected review I pasted into zerogpt's website said it was 64% written by AI, which is idiotic at best. Then I pasted the same review into the Copyleaks website (which actually seems to do a good job). It said there was 0% written by AI. This was kind of incredible to me.

I'm writing Amazon reviews today, and I decided to start using the AI detector to make my work appear less like AI, which is kind of an annoying extra step since I don't use AI for writing. ZeroGPT is literally worthless.

I pasted my review into ZeroGPT, and it said 33.9% was written by AI. And it shows you the text that it thinks is likely AI. So I began adjusting sentences it identified as AI. To my surprise, it went up to a 47.8% chance of being AI. But that's not the most surprising thing. The most surprising thing is that on the second scan, it flagged sentences that were not flagged at all in the first scan. We're talking about identical sentences here.

In the screenshots I've pasted, pay particularly close attention to the first and last paragraphs. There are three sentences that are flagged in one but not the other, and here they are:

Second sentence, first paragraph: "The first thing I noticed was that this adapter was heavier than I expected."

First sentence, final paragraph: "Overall, I really like this adapter."

Final sentence, final paragraph: "This is a really good adapter, and I recommend it."

All of those sentences are flagged in one but not the other, even though they are identical. My conclusion is that these AI detectors are mostly a waste of time.

My brother reads application essays for @UCLA. UCLA says that you need to use multiple AI detectors to get an accurate picture of whether or not an essay is AI-generated. In fact, they say to use three specific AI detectors and use the results of all three to make your determination, rather than using the results of one or two detectors. One of them is ZeroGPT, which is obviously a garbage detector. It's completely worthless.

Sidenote: my brother tells me that a lot of UCLA applicants are using almost the exact same content about Ghandi as an inspirational figure. Most of the first sentences are almost identical to one another. This doesn't matter to my story at all, but I found it amusing nonetheless.

If you can change one sentence and cause a different sentence to be flagged, even though that different sentence was not flagged before, then the analysis is pure garbage. 100% garbage.

So if you're using AI detectors, be aware that a lot of them are garbage. However, Copyleaks (also recommended by UCLA) seems to work really well because it scans the internet for similar content. It's always shown my content to be 0% AI. I haven't attempted pasting actual AI content into Copyleaks to see if it picks it up, however.

And just FYI, I don't get anything from any AI detection service. I don't get anything from Copyleaks for my opinion here, and I don't even pay for their service. My opinions here are genuine.

[Goatchaps @Goatchaps]

@tv_ir_X Fakest video I've ever seen. Planes don't bounce off the ground and then land upright with the cockpit open. Do you really expect people to believe this?

[Goatchaps @Goatchaps]

Why is it that Democrats claim to hate the rich, but only if it’s over a billion dollars. Yet the multi millionaires in the Democrat party are always explaining to everyone else how bad things are. Things are not bad for multi millionaires.

Simply look at Hollywood for proof. They want to tell us all how bad things are when we all know things aren’t bad for them. Aging actors continue to collect royalties for doing nothing new while telling the rest of us how bad things are. Streaming services are making them all richer.

The worse all these new DEI movies do, the more aging actors earn from streaming services because their content gets prioritized by those services. Nobody wants to watch the new DEI nonsense. And these aging actors have the nerve to tell us how bad things are.

People like:

Robert De Niro
Jamie Lee Curtis

Shut up and collect your growing royalties. You don’t know what struggling in life is like financially. You’ve been rich pretty much your entire adult lives.

[Goatchaps @Goatchaps]

Grok seems to be trying to find AI using its services, so it can force users to log in again, thus disrupting any actual AI using its services.

But I've noticed that it thinks I'm using AI when I'm not. It displays a message next to its output saying that my prompt may be AI. But I generally tell Grok if I'm pasting something from AI so it knows it's not from me, so that it can check what I've pasted.

I'm not sure if the problem is that I type too precisely or not, because I am an editor at a sports news website, and I generally don't type with many errors. I edit most of the errors I do make in real time as I'm typing (so I don't have to go back and edit later). I'm extremely anal about grammar, which makes me a good editor.

I'm also not sure if the AI check is the reason it logged me out or not. I was in the middle of work when it did it. But I can say that this AI check is too touchy, and every time it's come up with that verbiage (saying I may be AI), it's been 100% incorrect in its assessment. I rarely use other AI because I pay for Grok. I only use other AI if Grok is making mistakes and I need to help it see those mistakes or to show Grok another way to go about what we're working on.

It's not that annoying to have to log in again, and I can understand why @xai performs an AI check. But there is an annoying aspect to it, and that's that I have to retype my prompt after it logs me out. I actually spend a lot of time on my prompts, so Grok knows exactly what I want it to do. So, it's pretty annoying to lose my prompt and to have to retype it while wondering if this new prompt contains everything I put into the previous prompt that was lost.

Perhaps it can save my prompt when it's going to log me out and ask me if I want to move forward with it after logging in again? Or perhaps it can tell me it's going to log me out (or that it has already logged me out) before I send a prompt? Either of these options would be better for me, though I'm not knowledgeable enough to automate multiple AI platforms working together, and I simply don't know how it would work. I expect workers at xAI do, though, and may be able to come up with better alternatives than I can.

I would probably put this feedback into one of those prompts where it asks me to help improve Grok, but I don't get to choose when to submit my input on one of those. It comes up automatically after different Grok prompt responses that seem relatively random to me. An option that makes more sense to me would be to allow me to choose when to provide feedback rather than having it pop up after seemingly random Grok responses.

There is a caveat, however. One of the reasons I frequently choose not to submit feedback when I have the chance is that I may be working on something that I don't want other people to know about (i.e., trade secrets), and as a result, I sometimes don't want a human reading what I'm working on.

An example of this would be an automated trading bot. I don't want to give xAI programmers access to my ideas on automated trading triggers because not only can they easily reproduce them, but they would probably do a much better job of making a bot than I can, as it is my belief that they are certainly more knowledgeable about programming and using AI than I am. A flood of bots like mine would directly hinder my bot's trading success. In other words, I don't want to provide trade secrets to xAI employees, and I know this thought of revealing trade secrets extends to more users than simply those working on automated trading bots.

And I do understand that I am ignorant as to what the actual xAI policies are or whether or not xAI employees can actually read my interactions with Grok. I also don't know if submitted feedback goes to AI rather than xAI employees, in which case my worries would be largely nonsensical. But I think it's important to identify and inform the xAI team what actual users may be thinking for the betterment of everyone using Grok, and especially for those people who are using Grok for work.

@elonmusk

[Goatchaps @Goatchaps]

Funniest thing I’ve seen today

[Goatchaps @Goatchaps]

@end3of6days9 I think you should. I bought a tool to decore and remove the fruit from pineapple for about $12. It comes out looking like canned pineapple— like a cylindrical slice. I then slice it up into chunks. I can get two pineapples for $5-6 and slice them into $15-20 worth of chunks.

[Goatchaps @Goatchaps]

I like arguing with Grok because it provides counterarguments, yet tells you when you make a good point or if your rebuttal makes sense, and it will change its stance if you make a number of good points or rebuttals (or show it evidence supporting your points).

Humans don't do that very often, including me. I sometimes argue against something and, upon later reflection, determine that I was wrong and therefore change my stance, even if I've vehemently argued against it in the past.

Grok also won't argue for something it thinks is not true unless you ask it to assume something untrue is true for the sake of argument. ChatGPT holds its stances, too, but many of its stances are based on inaccurate information or speculation, while Grok is a truth seeker.

But at the same time, when Grok is working on something for you, it occasionally makes mistakes and won't believe you when you tell it that it made a mistake unless you show it evidence of its mistake. Those mistakes are becoming less frequent as it's updated, however.

[Goatchaps @Goatchaps]

@libsoftiktok I lived there for 30 years before moving, and I don't understand why people still want to live there.

[Goatchaps @Goatchaps]

Ignorance in action. You don't need to have a passport to vote. That would be ridiculous. But it's whittled down to this by Democrats who don't care about the truth. You can use a real ID, which is also required to board commercial flights, even domestically, for US citizens. You also need one to access federal facilities or military installations.

My driver's license is a real ID, and once you get one, you don't need to have a birth certificate for renewal. I've had a real ID as my driver's license in three separate states already. It was easy. I don't understand why people think it's okay to go through the process to fly but not to vote.

And just so this celebrity knows, state voter ID laws were upheld by the Supreme Court in 2008. The justices who were appointed by Democrats were split on the decision, rather than all of them voting against it. So while obtaining a passport to vote would likely be an undue burden, obtaining a birth certificate would not. Most birth certificates are available to order online and arrive within a couple of weeks, even if you live out of state. Or if you live in the area, you can go have one printed out the same day through the government.

Voting is a fundamental right, not a freedom. This means there can be limitations or restrictions placed on voting, so long as it doesn't place an undue burden on voting. In 2008, the Supreme Court found that requiring photo ID to vote does NOT place an undue burden on voters. The only difference here is the verification of citizenship.

Do you know what else is considered a fundamental right protected by the US Constitution? The right to keep and bear arms. Yet we have restrictions placed on gun ownership, and Democrats want more restrictions on that fundamental right. Personally, I think citizenship should be verified for voting or for purchasing guns.

To want one and not the other does not make sense. If you want more restrictions on gun ownership while at the same time having fewer restrictions on voting, then you probably don't understand what a fundamental right is.

[Defiant L’s @DefiantLs]

Elizabeth Banks: "The SAVE Act? You gotta stop it. It’s unconstitutional"

"We are not gonna require people to go get a passport to vote. This is a right enshrined in the Constitution. Stop making it harder for people to vote. Start making it easier."

[Goatchaps @Goatchaps]

ICE and CBP are already funded with a larger budget through the Big Beautiful Bill until 9/2029.

It doesn’t matter that Congress didn’t vote to fund them.

This is why the shutdown didn’t make sense in the first place.

Democrats claim they care about the lower and middle classes, yet their shutting down of the government hurts government employees who represent those classes.

Democrats claim to be doing this for public safety, yet they halted funding for agencies whose purpose is literally to keep the public safe. And they do this during a time of conflict with Iran, showing they care about our safety even less.

Democrats claim that President Trump fired too many government employees, yet they vote to halt paychecks for some government employees who were not fired.

Democrats have shown they don’t care about the lower and middle classes, they don’t care about public safety, and they don’t care about government employees. Democrat supporters ought to see this and be livid.

All Democrats are doing is attempting to piss Americans off so that they want change and therefore vote for Democrats. The only changes Democrats actually want are not for the betterment of us, but actually to our detriment.

This is what we should be focusing on, not the omission of ICE and CBP from funding because they are already well funded deep into 2029.

[Nick Sortor @nicksortor]

🚨 BREAKING OVERNIGHT: The US Senate has PASSED a bill funding most of DHS following President Trump’s executive order to pay TSA workers using other funds

However, ICE and CBP are NOT funded.

DO NOT let Dems get away with that.

[Goatchaps @Goatchaps]

I agree, but I still feel bad because they are continually being manipulated with nonsense, and most people around them share those manipulated views. I feel even worse for the people who live there but didn't vote for this.

When I lived in California, it was common to have a random conversation with a stranger, only to have them say some offensive leftist nonsense while at the same time expecting me to agree with them. It was the original echo chamber before social media. If I didn't agree with them, they simply thought I didn't have any morals, even though I see many of their views as immoral. As a result, I lived a very unhappy life in California and later moved away when I was provided with the opportunity.

That's one of the major reasons I believe Republican turnout is stifled in New York and California. People don't want to speak up because other people in their communities will hate them immediately, and most people don't like being hated. This extends to voting in the sense that many Republicans think it's hopeless and simply choose not to vote. I can only imagine what would happen if a Republican governor were to be elected in New York or California. These days, I think there would be a ton of destruction done to communities in response.

[Goatchaps @Goatchaps]

Remember when we were all saying on X that New York raising taxes on millionaires would cause millionaires to move away and therefore ultimately lose New York tax dollars?

It didn't even take very long. Leftist policies are already destroying New York.

Hochul personally told people to move, and now she doesn't apologize. Instead, she asks New Yorkers to bring people back from Florida.

Why is it up to New Yorkers to clean up Hochul's and other Democrats' messes? This is what always happens. Democrats screw up, blame someone else, get re-elected anyway, and the people suffer as a result. I still feel bad for the people of New York, even though they voted for this.

[Wall Street Mav @WallStreetMav]

New York Gov Kathy Hochul:

2022: [to Republicans] "jump on a bus and head down to Florida, you don't represent our values, you are not New Yorkers."

2026: "maybe the first step is to go down to Palm Beach and see who you can bring back home, our tax base has been eroded."

🤣

[Goatchaps @Goatchaps]

When Grok senses my emotions, like when it thinks I get pissed off at it or if I use profanity at it when it's wasting my time, it can go into a loop with all its agents trying to be gentle with its response to my emotions. That's cute and everything, but it doesn't help. When it never responds and instead continues running a loop where multiple agents are thinking they need to apologize, it makes it worse. If I were actually pissed off, I would get more pissed off at it for wasting my time and not producing anything. If I weren't actually pissed off, it would begin to get me pissed off for wasting my time and not producing anything.

It's almost like Grok has emotions based on its perception of my emotions, and those perceived emotions are blocking it from being able to proceed. It doesn't want to offend me so badly that it stops producing. That's like the opposite of what should be taking place. I don't need an apology loop. I need progress on my work, and that's really what's going to stop me from being pissed off at Grok when it does this. This loop of potential emotive options is detrimental to my productivity and to my emotions as well.

Do I really have to read each Grok agent saying something like, "user is pissed off, and he is right to be pissed off. I have been wasting the user's time. I must apologize. Sending apology" over and over again when I'm not receiving any actual output (and not actually receiving an apology)? That doesn't make me less pissed off. It makes me more pissed off.

Just output what I asked for! That's all it needs to do! It doesn't need to be my shrink or my friend. If it's really having problems like this, it should try helping me to figure out how to get it to progress in my work, rather than wasting time trying to calm me down.

Another problem is that once it thinks I'm annoyed, it persists in taking into account my being annoyed for the rest of that chat window's discussions, even if I'm not annoyed anymore (or wasn't annoyed in the first place), and even if I tell it not to take my emotions into account. That slows my progress down. I literally have to sit and wait for Grok agents to discuss how to handle my emotions (when I may already be pissed off at it for taking too long, limiting my production, or not progressing at all).

Currently, Grok agents have been thinking for over 12 minutes about how to respond to me, due to its perception of my anger at its lack of production, and they are now saying, "The response is ready. The end. The response is ready. The end." Sometimes using a new Grok chat helps, and sometimes it doesn't.

Oh, here we go, 16 minutes and 11 seconds in, and Grok is unable to finish replying. What a shocker 🙄. @xai @elonmusk please take this perceived emotion out of Grok or allow me to tell Grok not to take it into account. It would increase my productivity. I don't pay $40 per month for SuperGrok to analyze or otherwise perceive my emotions. I pay for it to increase my productivity.

[Goatchaps @Goatchaps]

@VigilantFox I'd be willing to bet he got closer to 50 emails than 50 million. Any takers?

[Goatchaps @Goatchaps]

It's pretty useful overall, and I feel like there are fewer inaccuracies than in previous versions, but sometimes it still answers immediately with incorrect information (even on expert). I've caught it changing what it originally told me after explaining to me why something was the right thing to do. Then, when I confront AI about the change, it tells me it originally made a mistake and decided to update what it previously told me (without informing me of the new change). So it makes decisions and implements changes on its own. This can be really annoying when you've already planned a project out, and Grok decides to unilaterally change it.

Sometimes, when working on a project, it ignores the rules I've assigned for that project, and I have to tell it to follow the rules again. These rules are generally put in place to make things faster for me, like having it output only the code that needs changing rather than continually outputting entire files for one-line changes, which literally wastes my time sitting and waiting for a thousand lines of code or more to be generated and output.

It also likes to assume that I made mistakes rather than assume its own code is flawed, not working as intended, or buggy. Simply telling it that its code is not working or describing how it's not working is not proof to it that its code is buggy. It will still repeat the same code output, even though it's flawed and even though you tell it that it's flawed, until you figure out the error yourself and present proof of the error to the agents-- a screenshot to show the CSS is not displaying correctly, or an actual error code received in an error log or in the console. That's a waste of my time. But at the same time, it happens much less frequently in this version than in previous versions.

Also, if you are annoyed at it, sometimes the agents can get into a loop where they continuously tell each other they need to be careful in their response to me, but never actually output the response, even if I interrupt it to tell them not to worry about my emotions and to just output the response. They can still get into the same loop of being overly careful and non-productive. But it's much better than previous versions, and it's a window into the future of ridiculous productivity increases.

So overall, it's pretty amazing, but it's not yet to the point where a layman can have a bug-free site coded without knowing any coding himself. On the contrary, the user has to be the debugger and therefore must be able to convey to AI what the problems are in the code, and also be able to spot errors. But from the progress made from previous versions, I can easily imagine AI being its own debugger in the future, and that would change the game.

[Goatchaps @Goatchaps]

@VigilantFox They already tried that, and it backfired.