215 get a cab phone number

My wife (30), who I (29) love dearly and have for nearly 12 years, has the itch and I do as well. We just bought a house, our jobs are steady (though we both plan to go back to school/ move around within the next year so we can get a bit more financially comfortable.) We could make it happen right now but we’re deciding to wait at least a year before trying anything.
I’m s******g pancakes about the whole idea, for a number of a reasons.. but that’s not the point of the post.
My questions is for those who were nervous about having a baby: What was something that pleasantly surprised you before, during, after child birth? i.e cute moments, feelings and emotions that were comforting or memorable, or made you say to yourself, “Oh, I get why people do this.” Lol
To elaborate, I’m a preschool teacher and I love kids - I can’t wait to be a dad because I’m going to kick ass at it (I know I’ll inevitably fuck something up along the way and I’m very aware of that!) But I have horrendous anxiety and conjure up horrible ideas in my head about future events 🙃🫠
So I’m looking for some support in things to look forward to when starting a new family. Any story you have to share is welcome 😊
Have a great week everyone!

What's up!
I have been posting some looking-for-friends-subs in the past, have deleted some of my post to start anew and now I thought I'd give this one last try. Yes, I am reading the various complains about not being able to find some friends on here and with most of them, I totally agree, but if you want to find friends, stop complaining and start doing something for that! By this point I just think it's hilarious seeing people posting this rant-posts here, knowing I have messaged those people and didn't even get a response! Take a wild guess why you are not getting any friends…
Anyways, my luck has been decent, and I am just tired of trying, so for one last time, since I still really like the concept, I wanted to give it a try. If it doesn't work, yeah, I'll stay by myself, I don't know. Maybe it's for the better.
My hobbies haven‘t changed, so I just get them out to see if you share them!
... writing. Been writing since I can remember. It started with smaller stories, no real plot or characters, you know, weird shit someone at 10 y.o. would write. But it has been going on. Been starting to prepare my first novel in 2012 and have been writing on it a bit, before completely starting over. So I really want to get past my damn writer‘s block so I can continue with that and finally be able to look for a publisher. Maybe I just have my head in the clouds, I don‘t know
... gaming. Like everyone these days, gaming has been my second favorite hobby since I can think of. I love action games, shooters, strategy, puzzle, co-op-games. I just never really been that much into PvP or competitive games (because I suck at them and after some time, my mood gets.… bad you know). But if you have something to suggest, please don‘t hesitate!
... photography. If I can get the chance, I take a picture of everything. I want to capture the moment. I got over a 1000 pictures of my dog, because from time to time, he can sit perfectly, you know. But mostly, I am into photographing landscapes!
... cooking. But only when I have the time. And I would say baking as well. I love trying out new recipes and figure something out I have never done before! What is your favorite meal? Do you have any nice recipe to share?
... watching TV shows / films / listening to music. Yes, maybe a cliché but I really have to have something on in the background. Right now for example, I am watching crappy German reality tv even tho I HATE IT, but I just need something in the background you know. But at least you can find a good tv show if you look for it. Tell me about your favorites (especially if it‘s horror which seems to be my favorite genre!)
... also I am too much into tech-y stuff, like phones for example, keeping up to date. And depending on my mood, I love history as well, especially Ancient Rome during Caesar's reign, or stories about catastrophes like Chernobyl or the sinking of the Gustloff. Just some side notes haha.
So that was a bit more detailed than before, just a bit about me now:
I used to study (laws to be exact) but since I have to pay rent and everything (even though studying in of itself is free) I had to drop out and now I am working (cliché) at a supermarket. A job I used to love, but since about 2 years, some things have changed so drastically, that I desperately want to quit. But I have to pay rent, get food, pay for my pets, so you know, for now, I am stuck.
A few months ago I basically broke off every contact with my so-called „friends“ that I had around here. You know, being the only one always asking for meeting and especially the only one keeping a conversation alive gets exhausting after some years, so I broke off everything. And who would have thought, they never tried to reach me again. So I guess it was the right decision. Since then, I have started to look for people all around the world because I don‘t mind talking in English, talking to anybody who wants to talk about literally anything. The only thing I want is that you know how to keep a conversation alive and don‘t respond within a month or what. In this time and age we live in, it‘s easier than ever before to maintain contact, keep up a conversation. There are different ways to do that.
Okay, I don‘t want to end this on a negative note, so please. If you have made it this far. Let‘s chat. I don‘t care who you are, what you do, just hit me up. No matter if you‘re a guy, a girl, if you like girls or guys, just talk to me and I am sure we will make up a good connection. And one last thing, I am honest about anything, so if you ask, I will answer you honestly, even if the answer may not be that easy to hear!
Hope to hear from you soon!

My boyfriend (24m) and I (23f) have been together for almost five years. Since the beginning of our relationship his female best friend has always been around (24f). We can call her Abby.
She was initially always very nice to me and we got on well. I’m not the kind of person who does confrontation when things go wrong and I was quite naive in the beginning, However, there would be little moments which would make me feel uneasy at times but I always pushed them to the back of my mind since it was never overwhelmingly negative. There were moments were she would talk about how everyone at school thought her and my boyfriend were together since “they are so close”. (They went to the same school which is how they met). She would also talk about my boyfriends ex girlfriend in front of me and make jokes about how “crazy” she is.
One day a fake account starting messaging me on Instagram about things they have done with my boyfriend, and that my boyfriend is only with me because “I’m good for the family”. And at the beginning of my relationship Abby actually told me that my boyfriend told her I’m the kind of girl that is perfect for his mum. My boyfriend knew about the fake account but shrugged it off and I didn’t mention any suspicions, I too eventually shrugged it off.
After some time the unease started to grow on me. He attended a dinner with her and another friend of theirs without inviting me, which I know it’s not really an issue but I was also friends with her, so I kinda thought it was strange. And my boyfriend did not invite me either. It had been a long time of us dating, perhaps over a year and half when I finally felt some confidence to speak to my boyfriend about the uneasiness.
I told him I don’t feel comfortable with him going without me and he got really mad at me and said he won’t go but he made me feel guilty for what I felt. I realised that if he didn’t think it was appropriate himself he wouldn’t have gone so why should I stop him? So I told him he should go and he did, knowing how I felt about it. Abby is very touchy feely with him so that was at the back of my mind and again, I am very non confrontational so I didn’t do anything about it which is also my fault. What hurt me most is how my boyfriend reacted when I expressed my concerns and still went.
Fast forward months later we had another arguement about it because I tried to express my concerns about the friendship again. And this time he got so mad at me that he messaged Abby telling her he’s cutting her off because of me but I didn’t ever ask him to cut off her at all. I tried to fix it by talking to Abby but she said “listen, (bf) tells me everything about you and we are best friends and there’s nothing you can do about it” I was confused about that statement but I was still so nice and told her I will fix everything and I apologised. Since then she never spoke to me and my boyfriend told me she was cutoff although I never told him to.
A few months later my family and I caught covid and we lost my father. I am the eldest of three siblings (youngest only 6) So I had a lot of responsibilities and I feel like I didn’t get to grieve properly With what followed.
My boyfriend and I were logged into on each other’s IG account. One day shortly after my fathers death I saw messages from a woman to my boyfriend which were quite affectionate. I met him to talk about it in person. He put his head low and didn’t say anything, he just had his head in his lap as if he didn’t know what to do but I still comforted him and politely asked to see his phone for the first time in our relationship. I didn’t want to ever be that person, but I recently lost my father and I had the mindset that nothing could make me feel worse than that feeling.
My heart sank when I saw the contents of his phone. Since the beginning of the relationship he told me how sending red hearts ❤️ to the opposite gender means more than friendship and it’s very intimate. This is because I once commented 😍 on a males artwork on Instagram (which was of a female render). And since then I never commented any hearts on any male page etc.
I saw lots of messages between him and girls he were friends with which were quite affectionate and in fact included these hearts he said I could not send to males. The most hurtful messages were between him and Abby. I went back to the night that everything blew up and he “cut her off”. He in fact sent the message explaining how he is cutting her off because of me, but instantly said to her “she’s going on and on crying to me about us” but I was upset because of his reaction and how he got raging mad when I tried to communicate my feelings. They were saying very horrible things about me including saying I’m ungrateful and Abby told him he should leave me. And he replied “Yeah I’ve given her enough chances” which hurt to my core, because I never did anything to hurt him. I don’t have any male friends and I’m very conservative. He had access to my social media from the beginning even before I saw any of his, because he had trust issues from his past relationship. They continued talking to each other long after that, saying every detail of our relationship including any arguements we had. And he had plenty of hearts ❤️ there too which only bothered me because he said it means more than friends etc.
I was devastated and was ready to leave. I tried walking away and he physically blocked my path to walk and wouldn’t let go. I went back to the car and he physically begged me not to leave him. I told him I will give him one more chance but have still since then been let down by finding other messages to girls.
He said he did all this because he “had a moment of weakness” and was going through a lot when I lost my father. Since then lots of time has past but I can’t seem to move forward and trust him. He still gets mad when I try to talk about it and I feel very alone. One time we had an arguement and he kept repeating how he “cut people off for me” which makes me feel terrible. I just want to move past it and stop feeling hurt about what he did to me. He says he was there for me when I lost my dad but I felt so lonely with everything he did to me at that time too.
Am I the AH and do you have any tips to get through this? Thank you and sorry if my English is bad.

The runecrafters of the Zamorakian Mage Institute have a problem: the Saradominist occupation of the essence mine makes it difficult to get the pure essence they need to keep up with the growing demand for runes. Many ZMI members are trying to get their hands on some extra essence by waylaying outsiders trying to use the Ourania Altar, but Herbert may have a better idea.
His travels in the more civilized parts of Gielinor have seen him acquainted with many adventurers with more pure essence than they can reasonably craft themselves and a voracious need for runes. He has decided to offer to sell these adventurers any quantity of runes, for the piecewise price of a few pure essence¹ and a crafting fee^2. The ZMI may even be willing to offer a discounted crafting fee for bulk orders^3, bringing it down to the current price of buying rune packs from shops.
1: At least 2 essence per rune ordered, potentially more essence per rune for higher-level runes. 2: Crafting fees start at 20% above the shop sell price of the rune. 3: An order can qualify for a discount both by total number of runes purchased, and for containing multiple rune types. Multiple types of runes get a discount because it's easier for the ZMI crafters to fill.

Large language models (LLMs) have received an increasing amount of attention from all corners. We are on the cusp of a revolution in computing, one that promises to democratize technology in ways few would have predicted just a few years ago. Despite the transformative nature of this technology, we know almost nothing about how they work. They also bring to the fore obscure philosophical questions such as can computational systems understand? At what point do they become sentient and become moral patients? The ongoing discussion surrounding LLMs and their relationship to AGI has left much to be desired. Much dismissive comments downplay the relevance of LLMs to these thorny philosophical issues. But this technology deserves careful analysis and argument, not dismissive sneers. This is my attempt at moving the discussion forward.
To motivate an in depth analysis of LLMs, I will briefly respond to some very common dismissive criticisms of autoregressive prediction models and show why they fail to demonstrate the irrelevance of this framework to the deep philosophical issues of in the field of AI. I will then consider the issues of whether this class of models can be said to understand and then discuss some of the implications of LLMs on human society.
"It's just matrix multiplication; it's just predicting the next token"
These reductive descriptions do not fully describe or characterize the space of behavior of these models, and so such descriptions cannot be used to dismiss the presence of high-level properties such as understanding or sentience.
It is a common fallacy to deduce the absence of high-level properties from a reductive view of a system's behavior. Being "inside" the system gives people far too much confidence that they know exactly what's going on. But low level knowledge of a system without sufficient holistic knowledge leads to bad intuitions and bad conclusions. Searle's Chinese room and Leibniz's mill thought experiments are past examples of this. Citing the the low level computational structure of LLMs is just a modern iteration. That LLMs consist of various matrix multiplications can no more tell us they aren't conscious than our neurons tell us we're not conscious.
The key idea people miss is that the massive computation involved in training these systems begets new behavioral patterns that weren't enumerated by the initial program statements. The behavior is not just a product of the computational structure specified in the source code, but an emergent dynamic that is unpredictable from an analysis of the initial rules. It is a common mistake to dismiss this emergent part of a system as carrying no informative or meaningful content. Just bracketing the model parameters as transparent and explanatorily insignificant is to miss a large part of the substance of the system.
Another common argument against the significance of LLMs is that they are just "stochastic parrots", i.e. regurgitating the training data in some from, perhaps with some trivial transformations applied. But it is a mistake to think that LLM's generating ability is constrained to simple transformations of the data they are trained on. Regurgitating data generally is not a good way to reduce the training loss, not when training doesn't involve training against multiple full rounds of training data. I don't know the current stats, but the initial GPT-3 training run got through less than half of a complete iteration of its massive training data.[1]
So with pure regurgitation not available, what it has to do is encode the data in such a way that makes predictions possible, i.e. predictive coding. This means modelling the data in a way that captures meaningful relationships among tokens so that prediction is a tractable computational problem. That is, the next word is sufficiently specified by features of the context and the accrued knowledge of how words, phrases, and concepts typically relate in the training corpus. LLMs discover deterministic computational dynamics such that the statistical properties of text seen during training are satisfied by the unfolding of the computation. This is essentially a synthesis, i.e. semantic compression, of the information contained in the training corpus. But it is this style of synthesis that gives LLMs all their emergent capabilities. Innovation to some extent is just novel combinations of existing units. LLMs are good at this as their model of language and structure allows it to essentially iterate over the space of meaningful combinations of words, selecting points in meaning-space as determined by the context or prompt.
Why think LLMs have understanding at all
Given that LLMs have a semantic compression of the training data, I claim that LLMs "understand" to a significant degree in some contexts. The term understanding is one of those polysemous words for which precise definitions tend to leave out important variants. But we can't set aside these important debates because of an inability to make certain terms precise. Instead, what we can do is be clear about how we are using the term and move forward with analysis.
To that end, we can define understanding as the capacity to engage appropriately with some structure in appropriate contexts. This definition captures the broadly instrumental flavor of descriptions involving understanding. I will argue that there are structures in LLMs that engage with concepts in a manner that demonstrates understanding.
As an example for the sake of argument, consider the ability of ChatGPT to construct poems that satisfy a wide range of criteria. There are no shortage of examples[2][3]. To begin with, first notice that the set of valid poems sit along a manifold in high dimensional space. A manifold is a generalization of the kind of everyday surfaces we are familiar with; surfaces with potentially very complex structure but that look "tame" or "flat" when you zoom in close enough. This tameness is important because it allows you to move from one point on the manifold to another without losing the property of the manifold in between.
Despite the tameness property, there generally is no simple function that can decide whether some point is on a manifold. Our poem-manifold is one such complex structure: there is no simple procedure to determine whether a given string of text is a valid poem. It follows that points on the poem-manifold are mostly not simple combinations of other points on the manifold (given two poems, interpolate between them will not generate poems). Further, we can take it as a given that the number of points on the manifold far surpass the examples of poems seen during training. Thus, when prompted to construct a poem following an arbitrary criteria, we can expect the target region of the manifold to largely be unrepresented by training data.
We want to characterize ChatGPT's impressive ability to construct poems. We can rule out simple combinations of poems previously seen. The fact that ChatGPT constructs passable poetry given arbitrary constraints implies that it can find unseen regions of the poem-manifold in accordance with the required constraints. This is straightforwardly an indication of generalizing from samples of poetry to a general concept of poetry. But still, some generalizations are better than others and neural networks have a habit of finding degenerate solutions to optimization problems. However, the quality and breadth of poetry given widely divergent criteria is an indication of whether the generalization is capturing our concept of poetry sufficiently well. From the many examples I have seen, I can only judge its general concept of poetry to well model the human concept.
So we can conclude that ChatGPT contains some structure that well models the human concept of poetry. Further, it engages meaningfully with this model in appropriate contexts as demonstrated by its ability to construct passable poems when prompted with widely divergent constraints. This satisfies the given definition of understanding.
The previous discussion is a single case of a more general issue studied in compositional semantics. There are an infinite number of valid sentences in a language that can be generated or understood by a finite substrate. It follows that there must be compositional semantics that determine the meaning of these sentences. That is, the meaning of the sentence must be a function of the meanings of the individual terms in the sentence. The grammar that captures valid sentences and the mapping from grammatical structure to semantics is somehow captured in the finite substrate. This grammar-semantics mechanism is the source of language competence and must exist in any system that displays competence with language. Yet, many resist the move from having a grammar-semantics mechanism to having the capacity to understand language. This is despite demonstrating linguistic competence in an expansive range of examples.
Why is it that people resist the claim that LLMs understand even when they respond competently to broad tests of knowledge and common sense? Why is the charge of mere simulation of intelligence so widespread? What is supposedly missing from the system that diminishes it to mere simulation? I believe the unstated premise of such arguments is that most people see understanding as a property of being, that is, autonomous existence. The computer system implementing the LLM, a collection of disparate units without a unified existence, is (the argument goes) not the proper target of the property of understanding. This is a short step from the claim that understanding is a property of sentient creatures. This latter claim finds much support in the historical debate surrounding artificial intelligence, most prominently expressed by Searle's Chinese room thought experiment.
The problem with the Chinese room at its core is the problem of attribution. We want to attribute properties like sentience or understanding to the "things" we are familiar with, and the only sufficient thing in the room is the man. But this intuition is misleading. The question to ask is what is responding when prompts are sent to the room. The responses are being generated by the algorithm reified into a causally efficacious process. Essentially, the reified algorithm implements a set of object-properties without objecthood. But a lack of objecthood has no consequences for the capacities or behaviors of the reified algorithm. Instead, the information dynamics entailed by the structure and function of the reified algorithm entails a conceptual unity (as opposed to a physical unity of properties affixed to an object). This conceptual unity is a virtual center-of-gravity onto which prompts are directed and from which responses are generated. This virtual objecthood then serves as the surrogate for attributions of understanding and such. It's so hard for people to see this as a live option because our cognitive makeup is such that we reason based on concrete, discrete entities. Considering extant properties without concrete entities to carry them is just an alien notion to most. But once we free ourselves of this unjustified constraint, we can see the possibilities that this notion of virtual objecthood grants. We can begin to make sense of such ideas as genuine understanding in purely computational artifacts.
Responding to some more objections to LLM understanding
A common argument against LLM understanding is that their failure modes are strange, so much so that we can't imagine an entity that genuinely models the world while having these kinds of failure modes. This argument rests on an unstated premise that the capacities that ground world modeling are different in kind to the capacities that ground token prediction. Thus when an LLM fails to accurately model and merely resorts to (badly) predicting the next token in a specific case, this demonstrates that they do not have the capacity for world modeling in any case. I will show the error in this argument by undermining the claim of a categorical difference between world modeling and token prediction. Specifically, I will argue that token prediction and world modeling are on a spectrum, and that token prediction converges towards modeling as quality of prediction increases.
To start, lets get clear on what it means to be a model. A model is some structure in which features of that structure correspond to features of some target system. In other words, a model is a kind of analogy: operations or transformations on the model can act as a stand in for operations or transformations on the target system. Modeling is critical to understanding because having a model--having an analogous structure embedded in your causal or cognitive dynamic--allows your behavior to maximally utilize a target system in achieving your objectives. Without such a model one cannot accurately predict the state of the external system while evaluating alternate actions and so one's behavior must be sub-optimal.
LLMs are, in the most reductive sense, processes that leverage the current context to predict the next token. But there is much more to be said about LLMs and how they work. LLMs can be viewed as markov processes, assigning probabilities to each word given the set of words in the current context. But this perspective has many limitations. One limitation is that LLMs are not intrinsically probabilistic. LLMs discover deterministic computational circuits such that the statistical properties of text seen during training are satisfied by the unfolding of the computation. We use LLMs to model a probability distribution over words, but this is an interpretation.
LLMs discover and record discrete associations between relevant features of the context. These features are then reused throughout the network as they are found to be relevant for prediction. These discrete associations are important because they factor in the generalizability of LLMs. The alternate extreme is simply treating the context as a single unit, an N-word tuple or a single string, and then counting occurrences of each subsequent word given this prefix. Such a simple algorithm lacks any insight into the internal structure of the context, and forgoes an ability to generalize to a different context that might share relevant internal features. LLMs learn the relevant internal structure and exploits it to generalize to novel contexts. This is the content of the self-attention matrix. Prediction, then, is constrained by these learned features; the more features learned, the more constraints are placed on the continuation, and the better the prediction.
The remaining question is whether this prediction framework can develop accurate models of the world given sufficient training data. We know that Transformers are universal approximators of sequence-to-sequence functions[4], and so any structure that can be encoded into a sequence-to-sequence map can be modeled by Transformer layers. As it turns out, any relational or quantitative data can be encoded in sequences of tokens. Natural language and digital representations are two powerful examples of such encodings. It follows that precise modeling is the consequence of a Transformer style prediction framework and large amounts of training data. The peculiar failure modes of LLMs, namely hallucinations and absurd mistakes, are due to the modeling framework degrading to underdetermined predictions because of insufficient data.
What this discussion demonstrates is that prediction and modeling are not categorically distinct capacities in LLMs, but exist on a continuum. So we cannot conclude that LLMs globally lack understanding given the many examples of unintuitive failures. These failures simply represent the model responding from different points along the prediction-modeling spectrum.

LLMs fail the most basic common sense tests. More disastrously, it fails to learn.

This is a common problem in how we evaluate these LLMs. We judge these models against the behavior and capacities of human agents and then dismiss them when they fail to replicate some trait that humans exhibit. But this is a mistake. The evolutionary history of humans is vastly different than the training regime of LLMs and so we should expect behaviors and capacities that diverge due to this divergent history. People often point to the fact that LLMs answer confidently despite being way off base. But this is due to the training regime that rewards guesses and punishes displays of incredulity. The training regime has serious implications for the behavior of the model that is orthogonal to questions of intelligence and understanding. We must evaluate them on their on terms.
Regarding learning specifically, this seems to be an orthogonal issue to intelligence or understanding. Besides, there's nothing about active learning that is in principle out of the reach of some descendant of these models. It's just that the current architectures do not support it.

LLMs take thousands of gigabytes of text and millions of hours of compute to talk like a mediocre college student

I'm not sure this argument really holds water when comparing apples to apples. Yes, LLMs take an absurd amount of data and compute to develop a passable competence in conversation. A big reason for this is that Transformers are general purpose circuit builders. The lack of strong inductive bias has the cost of requiring a huge amount of compute and data to discover useful information dynamics. But the human has a blueprint for a strong inductive bias that begets competence with only a few years of training. But when you include the billion years of "compute" that went into discovering the inductive biases encoded in our DNA, it's not clear at all which one is more sample efficient. Besides, this goes back to inappropriate expectations derived from our human experience. LLMs should be judged on their own merits.
Large language models are transformative to human society
It's becoming increasingly clear to me that the distinctive trait of humans that underpin our unique abilities over other species is our ability to wield information like a tool. Of course information is infused all through biology. But what sets us apart is that we have a command over information that allows us to intentionally deploy it in service to our goals. Further, this command is cumulative and seemingly unbounded.
What does it mean to wield information? In other words, what is the relevant space of operations on information that underlie the capacities that distinguish humans from other animals? To start, lets define information as configuration with an associated context. This is an uncommon definition for information, but it is useful because it makes explicit the essential role of context in the concept of information. Information without its proper context is impotent; it loses its ability to pick out the intended content, undermining its role in communication or action initiation. Information without context lacks its essential function, thus context is essential to the concept.
The value of information is that it provides a record of events or state such that the events or state can have relevance far removed in space and time from their source. A record of the outcome of some process allows the limitless dissemination of the outcome and with it the initiation of appropriate downstream effects. Humans wield information by selectively capturing and deploying information in accords with our needs. For example, we recognize the value of, say, sharp rocks, then copy and share the method for producing such rocks.
But a human's command of information isn't just a matter of learning and deploying it, we also have a unique ability to intentionally create it. At its most basic, information is created as the result of an iterative search process consisting of (1) variation of some substrate and (2) testing for suitability according to some criteria. Natural processes under the right context can engage in this sort of search process that begets new information. Evolution through natural selection being the definitive example.
Aside from natural processes, we can also understand computational processes as the other canonical example of information creating processes. But computational processes are distinctive among natural processes, they can be defined by their ability to stand in an analogical relationship to some external process. The result of the computational process then picks out the same information as the target process related by way of analogy. Thus computations can also provide relevance far removed in space and time from their analogical related process. Furthermore, the analogical target doesn't even have to exist; the command of computation allows one to peer into future or counterfactual states.
Thus we see the full command of information and computation is a superpower to an organism: it affords a connection to distant places and times, the future, as well as what isn't actual but merely possible. The human mind is thus a very special kind of computer. Abstract thought renders access to these modes of processing almost as effortlessly as we observe what is right in front of us. The mind is a marvelous mechanism, allowing on-demand construction of computational contexts in service to higher-order goals. The power of the mind is in wielding these computational artifacts to shape the world in our image.
But we are no longer the only autonomous entities with command over information. The history of computing is one of offloading an increasing amount of essential computational artifacts to autonomous systems. Computations are analogical processes unconstrained by the limitations of real physical processes. Thus we prefer to deploy autonomous computational processes wherever available. Still, such systems were limited by program construction and context. Each process being replaced by a program required a full understanding of the system being replaced such that the dynamic could be completely specified in the program code.
LLMs mark the beginning of a new revolution in autonomous program deployment. No longer must the program code be specified in advance of deployment. The program circuit is dynamically constructed by the LLM as it integrates the prompt with its internal representation of the world. The need for expertise with a system to interface with it is obviated; competence with natural language is enough. This has the potential to democratize computational power like nothing else that came before. It also means that computational expertise becomes nearly worthless. Much like the human computer prior to the advent of the electronic variety, the concept of programmer as a profession is coming to an end.
Aside from the implications for the profession of programming, there are serious philosophical implications of this view of LLMs that warrant exploration. The question of cognition in LLMs being chief among them. I talked about the human superpower being our command of information and computation. But the previous discussion shows real parallels between human cognition (understood as dynamic computations implemented by minds) and the power of LLMs. LLMs show sparse activations in generating output from a prompt, which can be understood as dynamically activating sub-networks based on context. A further emergent property is in-context learning, recognizing unique patterns in the input context and actively deploying that pattern during generation. This is, at the very least, the beginnings of on-demand construction of computational contexts.
Limitations of LLMs
To be sure, there are many limitations of current LLM architectures that keep them from approaching higher order cognitive abilities such as planning and self-monitoring. The main limitation has two aspects, the fixed feed-forward computational window. The fixed computational window limits the amount of resources it can deploy to solve a given generation task. Once the computational limit is reached, the next word prediction is taken as-is. This is part of the reason we see odd failure modes with these models, there is no graceful degradation and so partially complete predictions may seem very alien.
The other limitation of only feed-forward computations means the model has limited ability to monitor its generation for quality and is incapable of any kind of search over the space of candidate generations. To be sure, LLMs do sometimes show limited "metacognitive" ability, particularly when explicitly prompted for it.[5] But it is certainly limited compared to what is possible if the architecture had proper feedback connections.
The terrifying thing is that LLMs are just about the dumbest thing you can do with Transformers and they perform far beyond anyone's expectations. When people imagine AGI, they probably imagine some super complex, intricately arranged collection of many heterogeneous subsystems backed by decades of computer science and mathematical theory. But LLMs have completely demolished the idea that complex architectures are required for complex intelligent-seeming behavior. If LLMs are just about the dumbest thing we can do with Transformers, it is plausible that slightly less dumb architectures will reach AGI.
[1] https://arxiv.org/pdf/2005.14165.pdf (.44 epochs elapsed for Common Crawl)
[2] https://news.ycombinator.com/item?id=35195810
[3] https://twitter.com/tegmark/status/1636036714509615114
[4] https://arxiv.org/abs/1912.10077
[5] https://www.lesswrong.com/posts/ADwayvunaJqBLzawa/contra-hofstadter-on-gpt-3-nonsense

Fellow Mechwarriors,
I am in need of assistance. I recently got Mechwarrior Destiny, A Time of War, and Campaign Operations in physical format and devoted the last week to reading through all of them cover to cover. All of the books were great, but there's a substantial difference between destiny's action first streamlined approach versus AToW's sheer crunch and numbers. I have a group of friends who are interested in dipping their toes into the franchise and trying out a campaign set in the Battletech Universe, and they at least have rpg experience through D&D, gurps, and 40k tabletop rpgs we have done. I'm looking for help and tips for those who have run these systems before as I set out to create a campaign. While they have expressed interest in the logistics and company management that campaign operations adds to play, I dont quite want to scare anyone off by making them calculate escape Velocity for given planets and fuel usage past what the basics of a campaign would require. Think of it as kiddy pool depth; not too much physics and basic math homework for logistics and maintenance, but not so shallow that we're running it as rolling a d6 to see if your techs can repair all your stuff.
If any background will help then I'd like to run this as a campaign set between the 3rd and 4th Succession War(s), and let them run a merc group so that there's a bit of freedom to how they approach things. Plus I don't have to explain the clans, Kerensky, star league, etc past the first session primer.
Has anyone found a workable hybrid of the two systems that they and their groups found fun? How compatible is character creation in AToW's system, but using alpha strike rules for larger combat encounters? Is it better to use destiny or AToW for on foot actions and non combat sessions? Any house rules that make either system more enjoyable or streamlined? (For example I don't like how destiny has every weapon system in a mech that's linked become non functional when one weapon is damaged).
I apologize for the massive walls of text but I want to get more people into our little universe in an enjoyable way(and capitalize on its resurgence), and while all of the books like campaign ops have fantastic guides for including them in general play and alpha strike campaigns, I find there isn't a concrete guide to using supplementary books and converting them to Destiny or AToW. Any help is greatly appreciated.

I know it’s over and I’ll never see you again, I know it’s my fault for chasing you around, Yet I can’t move on from you,
I can’t explain what came over me, Feelings so strong it’s like I couldn’t stop, Which was odd because I barely knew you, But I really wanted to get to that place, Now I’m listening to sad songs, Feeling so sad I could cry on this phone, Thinking of you makes me feel broken and alone,
I can’t go back in time to fix this, You are gone and I’m going to miss you, Like a story with so much left to say, You flew out of my life and I wish you had stayed,
I want to talk to you and say I’m sorry But I can’t, I want to see your face but I don’t dare grace, The place I worked at with you, All I can do is write out my shame, Hope that letting all this out, Will help me to move on, From someone I never kissed, Who I never loved but will miss

So, I went to bed as usual but was so tired I forgot to close the closet door behind my bed. When I went to sleep it was a normal, dreamless sleep until I was transported to a cruise ship where I wasn't me, I was watching through someone else's eyes as I heard the words "follow the kite" and wouldn't you know it, there was a red kite just floating in front of me with no wind to propel it. I ended up following the kite as it somehow turned corners and eventually stopped at a old fashioned wall mounted telephone, cord and all, and there was a note pasted on the wall above the phone that said "dial SEX ASYLUM" and there were letters under the numbers on the phone, so I watched as the guy who I was watching through the eyes of dialed SEX ASYLUM. Suddenly "I" was teleported to a minimalistic, old bedroom watching a crazed girl in a white nightgown sitting on the edge of her bed slit her throat after putting on a black wig. After that I woke up, questioned the word itself because it was a strange combination of words, said "sex asylum" in my head while half awake and saw the image again.

Greetings dear friends
I have recently purchased 7hz salnotes zero and possess ATH M50x. I have iPhone and iPad Pro. I use Apple lightening to 3.5 mm dongle to listen music on iPhone. As iPad Pro doesn’t have a 3.5 mm dongle I am looking forward to get usb-c to 3.5 mm dongle. I am planning to get an android phone once my current iPhone gives up. When looking for possible options I came across 7hz -71, and it seems to be in my budget. I stream my music on Apple Music. My question is 7hz -71 an good option for the dongle I am looking for? Are there any good alternatives in the same price bracket ?
Thanks in advancer
Advice

This is so illogical but I am so, so stressed about going into my part-time job tomorrow.
I want to quit but I need the money. I'm a student full-time, so this is just on the side, but omg. It is such a horrible job. Basically, I'm an evening receptionist at a consulting place. However, the full-time girl (who is there 5 days a week, 9-5), quit two weeks ago. So now from 9-5 there is no one answering phones, etc.
I come in at 5pm to 30 messages and people calling screaming at me because "they've been calling all day and not getting an answer". There is always a pile of stuff to do because no one is doing it during the day. It doesn't technically matter because once my shift is over... it doesn't matter, but man I hate being yelled at.
Its also SUCH short shifts. Literally 3 hours. So I should just go and zone out and not take angry phone calls personally. BUT IM SO STRESSED. Please help :(

I (19M) have a week old daughter with someone who I took care of for 9 months and ensured their life was a piece of piss so they could rot in a bed for the entire duration and didn’t have to move to clean, eat, etc.
I cleaned their whole flat (which was squalor) so we could move from the North East to East Yorkshire where I knew nobody but they had tons of family offering endless support financially and physically.
3 days after she was born, my daughter and her other parent had to stay overnight in a hospital as she was jaundiced and a little underweight. In the meantime, I was at home cleaning the place up so they didn’t have to come home to a mess.
I was told that my partners mother was dropping by for a little on her way home from the hospital, turns out she was just coming to the house to kick me out of it. I got told to pack my stuff, hand her the phone she paid for and, as expected, I saw everything I’d worked for to make a family crumble in front of my eyes. My name wasn’t on the tenancy, I have no mode of transport, I had nowhere to go when I got back. I’m now effectively homeless- I ripped up the baby scan pictures mid breakdown while “packing my stuff” (I was too much of a mess to do that and I was given a fucking bag for life to do it in), anyways I walked downstairs with my bank card, a coat and the clothes on my back- I got drove instantly back to the north east.
The reasoning for this is that they felt the relationship has to come to an end , this is fine and not my main issue.
They’re refusing to put my name on the birth certificate. They said it’s partly due to the scan pictures, moreso because they just “don’t think it’s a good idea” but I don’t believe that considering they knew kicking me out would make me not be able to see my child. Which is true considering the situation I’m in.
I didn’t get to say goodbye to my child, that’s really haunting me. They just keep saying “she can make that decision once she’s old enough”. I’ve never once been a threat or risk to either her parent or her, I feel like I’m being treat like one.
What can I do to essentially get any form of custody? I’m lost. I feel like my life is ruined and I’m offered no sympathy from my ex or their family.

And everyone knows that patrolling the Mojave makes you wish for a nuclear winter...
I used to be the guy who lived in his phone. Between IRL friends, online friends, Kik groups, and a long-distance relationship, there was a period of a few years where it felt like my phone was never out of my hand. But over time, a lot of that stuff has faded or went away all together.
I welcomed it, at first. Trying to maintain so many conversations could get hectic and it was nice to not have to keep a power brick on me at all times to keep my phone fully charged to keep up with everything. But as time went on, I started to miss all the socializing and interaction. I don't think I want ALL of that back..but some would be nice.
I figured starting small is the way to go. So I'm on here today to find one good, new internet friend. Once I can tell if we click or not, I'm going to take this post down and go from there. But for now, the goal is clear: find my new Internet BFF
Could that be you? Only one way to find it. Shoot me a message and let's get chatting!

GTA 6 has been on the horizon for some time, and there have been rumors left and right about the upcoming game from Rockstar. Though the company has not announced nor confirmed anything yet, we could be getting closer to a launch of the much-anticipated game. Rockstar North, out of Edinburgh, is in charge of Grand Theft Auto development, and they are hiring a new game tester alongside other interesting positions.
Rockstar's website lists 18 different job positions currently available for applicants, including the rather interesting game tester role. Diving into the position, Rockstar is looking for people with "Knowledge of current generation games consoles and PCs" and who have "Competent gaming ability." Of course, this would all make sense for a gaming company; however, this is the division of Rockstar that produces the GTA series. Thus, we can reach a limited number of conclusions, including the fact that we could be close to at least a GTA 6 announcement.
jobs rockstar north hiring a game tester could mean we are inching closer to gta vi
Looking ahead this year, we will hit the tenth anniversary of the GTA 5 announcement, which could be good timing for the next game's announcement. While it may be a little cliche, it would be a great way to drum up hype around the dedicated GTA community. We could see a whole celebration of what has happened this decade with GTA, from the main campaign to GTA Online, and even smaller things like a thriving role-play ecosystem.
Whatever the case may be, it will be exciting to see if Rockstar North has anything in store for us later this year. Therefore, stay tuned to HotHardware as we keep an eye out for updates on any GTA 6 news hopefully coming soon.

The cards were previously used for mining, they do not come with the retail boxes unless it's specified. They might have some wear from normal use you can see the pictures for reference.
They are cleaned and repasted. They have been tested with furmark and 3dmark to ensure that everything working as it should.
If you don't see the price, please scroll to the right of the table.
Timestamps for other cards
​

Brand	Model	Price	Quantity	Available
ZOTAC	3080 Trinity	$499	4	4
MSI	3070TI GAMING X	$439	4	4
GIGABYTE	3070TI GAMING	$439	4	4
ASUS	3070TI TUF	$439	4	4
ZOTAC	3070TI TRINITY	$439	4	4
PNY	3070TI UPRISING	$399	4	4

​
If you would like to buy, please PM me your PayPal Email address and phone number (for shipping label) I will send you the invoice. Thank you
The card(s) will be shipped from Thailand with DHL express service (Trackable and signature require upon delivery), it usually take about 2-5 business days for delivery. Prices included shipping to the US, but I can ship worldwide. I will ship the card(s) within 1-2 business days.
I will ONLY ship to confirmed address on PayPal. Every serial number of the cards will be recorded.
If you have any question, feel free to send me a message (no chat)

Now that was just in the case of the factory, on the rarer chance that you are sent to "The House" I cannot reach you, and something much darker is waiting for you. But I can help you out with the few notes I have of the place.
​

1. When you are selected by the house, you will know because you have just woken up, you will not have any of your items on you except your phone and a small totem with be in your pocket, remember these as they will be needed to save your life in multiple scenarios
2. THE GOAL OF THE HOUSE IS NOT TO ESCAPE unless there under one circumstance, Refer to rule 25 for that scenario.
3. There will be someone on a bunk bed who is sleeping above you, her name is "Sylvia" she is a shadow entity who seems to have wolf ears, she has a scythe and tactical gear of sorts, she is very powerful and also your best friend, feel free to wake her up at any time, she can protect you, but the entities will be more vicious when she is around.
4. Rule 3 lied WAKE HER UP IMMEDIATELY FOR YOUR OWN LIFE'S SURVIVAL
5. DON'T TRUST RULE 4 IT IS THEM THEY WILL LIKELY KILL YOU
6. When you leave your bed room with or without Sylvia, you want to head to the kitchen, there you will find some food and some knives you can take either the food or the knives
7. Watch for Crawlers, they lurk above the ceiling and are trained to kill in any and every scenario, use your knife to kill them or ask Sylvia to, if she does not comply, it is a shapeshifter. THAT IS NOT SYLVIA. She will help the Crawler kill you.
8. If confused about the place, think about the house as HARD MODE for the factory, some of the same rules apply, mostly the ones that don't require you to use your guns
9. At Midnight, Sylvia will no longer be an ally for you, she will blankly roam around the house, and act very friendly if she sees you, but it is no longer Sylvia, her mind has been taken over by her emotionless counterpart "Kei"
10. If you are CAUGHT by "Kei" do not fear for it is not the end, she still has to act like your friend somewhat, so she will ask to play a game. Agreeing to the game will mean you have to give her the crystal orb in your pocket, this will make her much more powerful, declining the game will make her much more vicious, you will respawn in the same bedroom you started in where Sylvia was, It will take Kei 5 seconds to realize what happened and 15 more to get to where you are, use the vents to hop from place to place much quicker.
11. Keep in mind that this place is as big as a mansion, you have many places to run and even more to hide, but if you are cornered, pray that she only wants to play "The Game"
12. When she mentions the game do not ask for the rules, because the only one is to survive however you can. If you do ask... Please, refer to rule 15.
13. The game is hide and seek, you can easily move places but the vents will have amplified sound, so they will know the general area you are going to
14. Kei is emotionless so she will not care about your cries of pain or your pleading, any emotion you show is grounds for her to become more vicious in her mind, you are annoying to her.
15. Kei will explain the rules and it becomes 10 truths and 5 lies, the game will also change from hide-and-seek to a game only she can tell you about, you then become out of my reach and I am sorry.
16. At 3AM Sylvia's body will have completely changed into Kei's, she will become less and less friendly slowly as she is obviously faking these emotions up until 3AM at that point in time, she's fed up with the game of cat and mouse and will do anything to kill you
17. at 3:30AM The house will start flashing random colors. Red means she's getting hotter, or closer to you... Blue means she's getting colder, green means she is too far away for them to tell, and yellow means she has given up.
18. If the lights ever stop flashing, Kei is right behind you. Pray, but not to your lord and savior, pray to Kei herself, she may find this amusing, as at least you are not using your annoyingly high-pitched voice.
19. The lights only ever flash green because the house has expanded from a big home into an ACTUAL Mansion, the cameras can only see so much...
20. Congratulations it is now 2:50AM and you now only have to worry about certain scenarios at certain times. Pray to Jesus at this time, Pray to Kei if you feel like it, she may change her behavior, positively or negatively is up to her, but praying to her will lean more towards positively. Failing to at least pray to Jesus will make this place Un-holy ground, any demon will be able to help Kei hunt you down.
21. You are free to leave the mansion at any time from after 3-4 AM, but depending on what time you leave after 3AM... you will need to face an opponent
    1. If you leave at 3:01-3:15 you will face the one named "Sensei" he is the one who taught "Team Shadow" which is a group that Sylvia and Kei belong to how to fight and ancient fighting techniques, although Kei likes to stick to distortion.
    2. If you leave at 3:16-3:29 you will fight Sylvia's brother... Mo. He is powerful but not as powerful as Sylvia, and he is powerful in a different way then Sylvia, he is also much more disciplined, and paid the most attention in Sensei's class.
    3. YOU MAY NOT LEAVE AT 3:30-3:35 Doing so will alert Kei to your position, but it won't matter because you'll already be falling through an endless void.
    4. If you leave at 3:36-3:59 it will be the leader of Team Shadow... His name is Cole though something will be off about him, his body parts will be shredded or scratched. he will be badly bruised and in a lot of pain, this is because everyone but Sensei on team shadow hates him, including Sylvia and Kei. Despite this advantage, make no mistake, he will still be the toughest opponent you've faced in your life.
    5. In the event you do not leave before 4AM the doors will close and you will need to wait for another opportunity to escape, Kei will take note that you passed up the chance to leave and one of two things will happen: Possibility 1: She will turn back into Sylvia and try to trick you, despite this it is technically a mercy. Possibility 2: She will become even MORE vicious and faster, also the walls will move inward by two rooms from each direction, shortening the area needed to find you.
22. At 5:00AM you will be teleported into an infinite hallway, Kei will be a few miles behind you. ***************RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN RUN***************
23. Make sure to run until your legs give out, she will catch you in a few hours, when she catches up one of two things will happen, she will jump-scare and then kill you or you will be teleported into an arena where you must fight her, you are given the choice of a sword, a scythe, or magma boxing gloves (That don't burn your hands) Whichever one you pick will also teach you a different fighting style
    1. The boxing gloves are... Never mind you already know. And if you somehow don't then it isn't my frickin' problem
    2. If you pick the sword, you will know how to fence, among other sword fighting styles.
    3. If you pick the scythe, you will know how to fight like Kei does, be better at her art then her.
24. If you win you are free to leave. She will bow and let you go, make sure you haven't killed Kei. If you somehow do, the entities will rush in and kill you, and Sylvia will emerge from Kei's ashes to help finish you off as well.
25. This will explain the rare scenario that Cole is there in the bed above you instead of Sylvia, the same rules will apply but the goal is now To Escape not survive.
    1. The house will turn a labyrinth and you had never seen Sylvia as Sylvia, she is just Kei. She attacks you because she thinks you are on Cole's side, you can try to side with Kei to kill Cole but then you Betray Cole, and she might not even believe you, if she doesn't then you have NO FRIENDS and you will most likely die.
    2. Death is only guaranteed in this scenario where the house turns into a labyrinth.
    3. The good thing is that it will only transform once Cole has woken up so you still have to follow rules as if it was a Sylvia scenario until you wake Cole up.
    4. The Ratio of Sylvia Scenario To Cole Scenario Is 98%-2% in the Sylvia Scenario's Favor
26. If you escape with Cole or Defeat Kei, then you will be here, just like when you escape the factory, I will be allowed to let you go back into the real world.

​
I really really do apologize for this inconvenience but IT is making me do these things, now run along, you have a factory to escape, or a house to survive in. Good-luck. -Fellow Inmate

Hi y’all! I’m new here but I am so thankful to have found this sub - it’s hard not knowing what to expect but it’s comforting to know I’m not alone. I found one spot about 1.5 inches in diameter in January 2023. I now have about 10 spots of differing sizes. Most about dime or quarter size. The first big spot has grown in size but also has some regrowth. Thankfully almost all of my spots are hidden by my long hair. I started with steroid shots right away but due to the number of new spots, my derm now has me taking Prednisone (1st week 60mg, 2nd 40mg, 3rd 20mg, 4th 10mg). I’m at the beginning of my 3rd week of oral steroids.
When my hair was about to fall/falling the spot would get super itchy and have a dull ache. During the 2nd week of Prednisone, those feelings stopped! (Keeping my fingers crossed) For those of you with regrowth, I was wondering what did regrowth feel like? Was it sore? Did it feel similar or different than losing hair?
I hope you all are having a great day!

Just fell for a scam website on reddit. I am subscribed to ufos and saw a post with a poster of Aliens and Van Gough's starry night. I thought the poster was pretty rad and ordered one.
The poster was listed at $24usd (a bit high but I liked the poster) and $5usd for shipping. After entering my credit card information and making the purchase I was charged for $38usd ($8 more than what I was quoted prior to making the purchase.)
After finding that odd I commented under the posted link on the reddit page, and was informed by another user that it was infact a scam website. Looking up the website it is fairly new and has stolen art and is very similar to other websites hosting the same "merchandise." The server it is on also hosts other websites with pretty terrible ratings.
I cancelled my card immediately and will be disputing the charge when it is done pending.
Anyways, don't be like me, do some digging before an impulse purchase.
I also filled out my address (for shipping purchases) and my phone number and email address. If there are other steps I need to take to protect myself please let me know in the comments.

I downloaded Stanford Alpaca and the 13B model and tried to see what it could make in Go. I used a n_predict value of 100000.
First test:
### Instruction: >write a golang program that calculates all the prime numbers up to n.
### Response: // Go Program for Prime Numbers 1 <= N < Integer Maximum // package main import ( "fmt" ) func isPrime(n int) bool { return true } /* This function returns always 'true' */ var i, j = n - 2; if ((i % j == 0)) && !isPrime(j+1){return false} for ( ; i <= j + 1; ) { // Check all numbers from one to the maximum integer value } return true
The result were not very good.
​
Second test:
### Instruction: >Write a golang program that queries google for the current time.
### Response: func CurrentTime() string { //Query Google and return result as String type url := "https://www.google.com/search?q=current+time" //URL to query from client := http.Client{Timeout: time.Second * 5} //Set timeout for HTTP request resp, err := client.Get(url) //Make GET Request with given URL and return Response & Errors if any if (err != nil){ //Check error response from server to determine success of the query fmt.Println("Error:", err); //Return error message in case there is an issue while making request }else{ //If no errors, return result as a string type respBody := strings.NewReplacer(resp) //Replace the body of response with empty String fmt.Println("Response:", resp); //Return Response message in case there is success while making request } //If no errors, return result as a string type }
Well it does look like something a programmer could base their work on.
​
Third test:
### Instruction: >Write a golang program that queries coinbase.com for the price of bitcoin.
### Response: func queryCoinBase(url string) { //Query Coin Base Website For Bitcoins Price //URL = https://api.coingecko.com/api/v3/currencies?vs\_currency=USD&limit=10 var resp *http.Response var client http.Client resp, err := client.Get(url) //Make a GET Request to Coinbase Website //Check for Errors and Respond Appropriately if any occur } if (err != nil){ return "Error Occurred" } defer resp.Body.Close() //Closes the Response Body }
I do think it is neat that the A.I. can write some Go code. I just needs a lot more work. The good news is our jobs are safe for now.

Hi guys! Been married about two years and I noticed that whenever me and my spouse are apart he will never call me on the phone unless I ask him to. He will only send update texts on what he’s up to. I get people are busy and have lives but it bothers me and it makes me feel forgotten about sometimes. Men do you guys not like calling your wives unless it’s a mandatory thing or it’s super important? Just trying to understand from the male mind.

Hi! I'm working on a project and I would like to calculate an approximation of the required battery capacity of the device. I came across this technical documentation from the semtech website. I wanted to verify the formula using the example they posted before proceeding, but the values I get are nowhere near the result they provided. I was wondering if I'm overlooking really something simple. What am I doing wrong here? I'm already banging my head out of frustration lol.
N = Number of times the device will be active per day
cU = Capacity after taking into account discharge rate (mAh) = C*(U/100)
T = Milliseconds per hour = 3600000
cA = Active capacity usage per day = N*(Ttx*Itx + Trx*Irx + Tmeas*Imeas)/T
cS = Sleep capacity usage per day = Is*( 24-((Ttx + Trx + Tmeas)*N/T) )
cT = Total capacity consumed per day = cA+cS
​
Given values in the example:

• Minimum current/sleep current: 5uA
• Maximum current/Tx current: 135mA
• Times the device will be active per day: 36
• Maximum current/Tx current: 135 mA
• Tx time/ToA: 250ms
• Rx time: 131ms
• Rx current: 13mA
• Measurement current: ~10mA
• Measurement time: 10ms

This is how I plugged in the values:
cA = 36*((250ms*135mA) + (131ms*13mA) + (10ms*10mA)) / 3600000
cS = 5uA*( 24 - ((250ms + 131ms + 10ms)*36 / 3600000))
cT = cA + Cs
In my understanding we're supposed to get around 1091mAh after this? or am I using the formulas wrong?