Date: July 31st, 2026 11:19 AM Author: spectacular station macaca
Imagine you need to run a dozen prompts in a row or in parallel using dynamic workflows and multiple tools/MCPs with gated verification steps propagating to shared markdown files all flowing downstream to a single deliverable
You could just drop the entire playbook into codex and ask it to one shot the whole thing while you take a shit
Date: July 31st, 2026 11:45 AM Author: light idiot plaza
It's no harder than using the regular gpt really. It does everything for you. It used to be just for coding but you can use it for anything now. Main difference is it will make a directory on your harddrive and generate a bunch of artifacts as it works, and it runs for up to 2 hours and you can delegate sub-agents to do different tasks.
Date: July 31st, 2026 11:14 AM Author: spectacular station macaca
When they update the system prompt for a new model the old ones perform sub-optimally
The only way to get anything close to consistency is harnessing multiple frontier models with Chinese open weights and intelligently designing a canonical repository as the gospel of truth from which everything flows downstream
Date: July 31st, 2026 11:31 AM Author: Flatulent Bateful Preventive Strike Bbw
I've never once used Claude, Gemini, or Copilot. I'm a ChatGPT man. The original is still the best. Although I don't really know because I won't branch out.
Date: July 31st, 2026 12:46 PM Author: Tantric bawdyhouse
believe it or not some jobs consist of using your brain and actually understanding the business they are running vs. a computer that objectively doesn't
and fwiw, the 'email girls' of the world are all in on this chatGDP shit and use it to write all their emails without any sense of irony
Date: July 31st, 2026 1:19 PM Author: gunneratttt (gunneratttt)
First you attacked other people for needing the tool, then you shiffed to concern about kids developing skills, and now you're arguing the tool is bad at what it does.
All of these things conflict with each other. You're just a grumpy old man and searching for a justification for your outrage over things being different than what you're used to.
Date: July 31st, 2026 1:59 PM Author: Tantric bawdyhouse
ppl currently needing the tool = ppl too dumb to write their own emails or codemonkeys not self aware enough they're ruining their own career
kids not developing critical thought skills = real problem that's already being seen in schools/colleges
bad at what it does = it seems like it does shittier but faster lower effort versions of tasks that you need to give it "prompts" to do. to the point right above this one, if there's no one with enough knowledge to give prompts then the prompts will be shit and we get in idiocracy
it's not being "grumpy old man" or any shit like that. it's more like identifying that this is objectively bad.
Date: July 31st, 2026 12:48 PM Author: Tantric bawdyhouse
kids being taught that they don't need to learn how to think is a real problem.
like for example, an AI bot probably will tell you that avg global temperatures rising are a relevant metric and completely validates everything climate scientists have been putting out for the last 30 years and AI dependent npcs will lap it up. people who actually can think can figure out in a few seconds why it's a bullshit metric.
Date: July 31st, 2026 12:58 PM Author: gunneratttt (gunneratttt)
But you're criticizing people that do know how to write emails for using a tool that can make it more efficient. No different than a boomer scoffing at email and chalking it up to people being too lazy to send a letter, and before that using a typewriter instead of hand writing.
You use all sorts of computerized tools to do your job. The only reason you feel differently about AI is because you're old. You grew up with email and wordprocessors so that's acceptable to you, but a new tool isn't, but only because it's new to you and so "that's not how you're supposed to donit!" And thats no different than every other old timer throughout history who rejects every innovation that occurs after 40.
Date: July 31st, 2026 1:22 PM Author: Arousing pea-brained locus
lol you sound like a child projecting his mental image of a "businessman making business decisions". also 180 that you attack people for "shouldn't you be doing something else right now?" without a hint of irony
Date: July 31st, 2026 12:56 PM Author: Tantric bawdyhouse
lol yep
AI Overview Average global temperatures are rising due to human activities that release greenhouse gases, primarily driven by burning fossil fuels, cutting down forests, and industrial farming. Earth's average surface temperature has increased by about 2°F (1.1°C to 1.3°C) since pre-industrial times, with the fastest warming happening in recent decades.To solve or slow down global warming, the world must reduce heat-trapping gas emissions and transition away from carbon-heavy practices. You can learn more about actionable paths via the US Climate Resilience Toolkit or NOAA Climate.gov.Key Solutions to Climate ChangeRenewable Energy: Switch from coal, oil, and gas to clean power sources like solar, wind, and water energy.Clean Transportation: Use electric vehicles, ride bikes, walk, or choose public mass transit instead of driving gas-powered cars.Protect Nature: Stop cutting down forests and plant new trees to naturally absorb extra carbon dioxide from the air.Save Energy: Use better insulation in buildings and upgrade to energy-saving appliances to use less power overall.
Date: July 31st, 2026 1:21 PM Author: light idiot plaza
In your own words, can you explain where in this paragraph it says that every climate model is correct, every projection from the last 30 years was accurate, every proposed remedy is effective or that disagreement with any climate scientist is irrational?
Date: July 31st, 2026 1:26 PM Author: light idiot plaza
His stated objection is that average global temperature is a “bullshit metric,” but this summary doesn't use temperature alone as proof. Or is his argument that global mean temperature is completely physically meaningless?
His point if I understand correctly is that ChatGPT is making people dumber because they're not thinking for themselves and it slips in a lot of baseline lib assumptions about the world that stupid people lack the tools to question.
Date: July 31st, 2026 1:29 PM Author: light idiot plaza
Yeah and my point is that the example doesn't show that. It's a compressed overview of the most basic science on the issue. Nowhere does it endorse every lib policy on climate.
Date: July 31st, 2026 1:35 PM Author: light idiot plaza
Well it's trivially true that we are absolutely wrecking the planet with our current activities and have done irreversible damage. Now what should be done about it as policy is what is up for debate. It lists some of the more kosher methods for fixing it that libs like to talk about. But obviously what really needs to be done is we need to kill off massive parts of the human population and force other countries to change their ways too. The problem with libs is they seem to think the west has to bear the weight for all of humanity. We have already spent so much money on this, and it has had little effect on reversing it.
Date: July 31st, 2026 1:44 PM Author: light idiot plaza
Yeah but that's because of its guardrails because of the PR blowback on companies. This is why they spend millions on "human ai alignment". Here is the thing though you can back GPT into a corner on these issues and read in between the lines.
Date: July 31st, 2026 1:48 PM Author: gunneratttt (gunneratttt)
It's a window into the lib mindset. They're more concerned with "doing something" to address a problem that we're responsible for even if the solution isn't calculated to have any impact on the problem. All the white guilt stuff is the same phenomenon.
The only way to address climate change is if the world collectively reduces emissions. But any country that doesn't will have a huge competitive advantage, so it would have tonhe enforced milarily. Or reductions could happen organically because of depopulation like you said. Developing countries would resist this enforcement and they're right to: the west industrialized off cheap energy and it's unfair that we'd keep our infrastructure and demand they remain poor. So we'd either have to pay their green industrialization or destroy them if they won't comply.
Any of this happening is pure fantasy. We're currently struggling in a war against a fairly weak country in large part because we need them to continue fossil fuels flowing or our economy will suffer. I think everyone has accepted this now which is why there's far less emphasis on climate policy than there used to be.
Date: July 31st, 2026 2:24 PM Author: light idiot plaza
Agreed. I think the honest assessment is that nothing is really going to be done about it fully until we have the technology to reach 0 emissions and even then it will probably be certain countries paying to make sure the entire globe has the tech. Up until then we will probably continue to pay unevenly toward minimizing it, and it will have minimal effect. Probably by 2100 there will be like a 3 degrees celsius rise and don't see that being stopped. Humanity can survive that, but not without massive disruption. One thing I'm certain about is that even if humans continue to burn fossil fuels into 2400 it won't wipe all life on Earth out (will wipe humans out though). If life survived the siberian traps it can survive us.
Date: July 31st, 2026 2:36 PM Author: gunneratttt (gunneratttt)
Our thinking is completely aligned on this.
There are way too many factors to predict with any confidence how this'll turn out after we're dead. Perhaps huge disruption will result in massive human collectivization that's unthinkable today. Or, more likely from my POV, it'll result in global war over increasingly scarce resources. Or maybe the same post-WW2 MAD Mexican standoff will hold out and we'll just die off slowly. Or maybe technology will be developed that we can't conceive of just like how no one in 1826 accurately predicted what 2026 is like.
For me, I put about as much thought into it day to day as I do the heat death of the universe. The consequences will happen long after I'm dead and I have zero impact on it in any case. I try to prioritize energy and concerns toward things I can impact.
Date: July 31st, 2026 3:23 PM Author: light idiot plaza
The thing I don't get about people denying it COMPLETELY, not just denying whether we have the right policy or whether the climate modeling has the right assumptions for long term projections is that it isn't just one institution with one set of incentives gatekeeping the data and measurements. It rests on a redundant measurement system worldwide with independent data sources, calibration procedures, and physical observables. There are multiple converging lines of evidence too. For instance the temperature is lining up with isotopic ratio and spectroscopy results for carbon and also what the laws of thermodynamics and conservation of energy would predict. You can't just fake or accidentally produce that for ideological reasons.
like 90% of climate scientists in the world are from places like Finland and Germany which have a massive degree of left wing ideological capture deeply rooted in their culture. For example Sweden has been a Socialist Country for over a century and there are Communist meeting houses in every city and they sent aid to the Vietcong and they're on the vanguard of allowing migrants into their country and causing massive societal upheaval as quickly as possible. Now how much 'stock' would you put into the median Swedish climate scientist given how wrong they are about everything else? this is a culture that even invented new words for pronouns for trans people lol. How fastidious do you think their studies really are on climate science?
Remember the head climate scientist at MIT said the Earth is warming up very slightly but it's normal and nothing bad will happen. Instead of refuting him they just said he's a crank and no one takes him seriously. But not "all" climate scientists believe in climate alarmism.
Date: July 31st, 2026 3:39 PM Author: gunneratttt (gunneratttt)
I think it stems from essentially the same psychology that results in the lib guilt thing I discussed above.
In the libs case they want to assuage their guilt about climate change by "doing something", even if it's completely performative. They are motivated by feeling they are good people regardless of whether that addresses the issue.
Likewise, reptiles deny the problem exists at all to avoid feelings of guilt and futility over an unsolvable problem. Their motivation is about relieving their own cognitive burdens.
The reason I think this is because how emotional and stubborn they are. A reasonable person would at least concede that they aren't scientists, have no ability to independently investigate this beyond reading other people's research, and that the weight of the evidence, at the *very* least, forces one to concede that anthropogenic climate change is a possibility.
That's why I focused in on something is indisputable: global average temperatures rising. If you can't get someone to admit that, then that person has abandoned rationality, hence there must be some emotional reason what the believe, particularly because of how strongly held it is. Attempting to reengage on it highlights this as well. They are emotionally invested in their positions being correct, but a rational person would be ambivalent since it's just a matter of objective fact.
Date: July 31st, 2026 2:45 PM Author: Tantric bawdyhouse
that's a lot of words and histrionic doomsday predictions without any explanation why it's even a short term problem let alone one that will wipe out humanity by 2400
Date: July 31st, 2026 3:02 PM Author: light idiot plaza
Are you saying you don't believe in anthropocentric climate change at all and that it has had no measurable effects and will not continue to have effects at all? Because that is completely physically indefensible. Obviously projections out to 2100 or further are model-dependent, and models have assumptions. But in that case state your better model than the ones done by places like NASA and Stanford
it's obviously the effect of oversampling from cities growing and too many thermometers. like all the reduction analysis has shown that the increase in temp is universally in major metros at night. just think if the temp at night where you live is 65 F now 200 years ago it was 63 F is the Earth really getting hotter or did they just build a ton of houses in the neighborhood and human activity traps a little heat around?
also the satellite data shows that more heat escapes the Earth when the Earth gets hotter than previously thought. the 'green house effect' is pure propaganda. More energy = particles have faster velocity = more shit escapes the Earth.
The Earth actually isn't meaningfully warming up and it's not tied to carbon emissions. Also the worst weather events on record happened in the 1930s and no one can explain that. Shouldn't we be seeing more adverse weather? and all the evidence points to no, we're not lol
Date: July 31st, 2026 4:27 PM Author: light idiot plaza
Humans not turning down money is a sociological argument that can explain what types of questions get asked, and why certain types of modeling and research get funded and institutionally selected for, but it can't explain instrument data from satellites measuring earth's radiation, surface and subsurface thermometers measuring temperatures of the land and ocean, spectrometers, co2 calculations etc. All of NASA and other organizations raw data is directly downloadable through their apis. I never said I don't think the predictive modeling could have institutional and political bias attached to it, but you have been making stronger claims than that, basically denying that temps have risen at all since the 19th century and that it isn't caused by increased carbon etc.
Date: July 31st, 2026 4:41 PM Author: Tantric bawdyhouse
not going this chat gdp slop word by word, but no, i didn't say that I don't think temps are rising. i'm saying that climate alarmists have never effectively communicated why it matters.
Date: July 31st, 2026 4:53 PM Author: light idiot plaza
So is what you are saying now that you accept warming but think advocates have communicated the consequences poorly? I would consider that a pretty significant walkback from your earlier argument.
Date: July 31st, 2026 4:46 PM Author: gunneratttt (gunneratttt)
There's also so much data collected by ordinary people that confirm it.
For example, maple syrup producers have been recording when mountain snowpack has thawed for hundreds of years. To reject climate trends you also must believe that the recorded secondary effects of it are also faked. That some agent is going around convincing ordinary people to misrecord something everyone plainly observe.
Date: July 31st, 2026 5:00 PM Author: light idiot plaza
It seems you imagine warming as a small uniform increase in comfortable ambient temperature. The Earth system is not one object with a comfort setting thermostat that you can just shift up a few degrees without catastrophic consequences.
Date: July 31st, 2026 5:04 PM Author: gunneratttt (gunneratttt)
But didn't you see his Google map link above? Some parts of Canada are frozen even in May! Sure, maybe all the temperate regions will become deserts or submerged, but that land may become arabale! It's a wash or maybe even a good thing!
Date: July 31st, 2026 10:58 PM Author: Tantric bawdyhouse
it’s entirely possible that it might be a good thing. A shitload of land on this planet is a permanently frozen desert useless for anything except oil extraction.
Date: July 31st, 2026 4:31 PM Author: gunneratttt (gunneratttt)
You routinely try to focus on reasons some may be biased (which no one disputes), because you don't have the ammo to tackle the merits. Ultimately it doesn't matter what their motives are. We're talking about analyzing objective facts. Whether they are pure-hearted or evil doesn't change reality.
You retreat to these things because climate is something everyone experiences and can collect data on. It's not something that's possible to fake.
Date: July 31st, 2026 4:38 PM Author: Tantric bawdyhouse
it's possible to fake/groupthink to death the concept that it will result in humanity going extinct in 300 years in order to secure funding at prestigious universities and/or elite govt institutions and to fit in with elite scientific circles in general.
Date: July 31st, 2026 4:53 PM Author: gunneratttt (gunneratttt)
The conspirators would have to be superhuman smart and capable to do that, which is a lot of work to secure university grants. It's a wonder why they didn't just go work at tech companies for 10000x the payoff with just a tiny fraction of the effort and no risk of getting caught concocting a global climate conspiracy.
Date: July 31st, 2026 5:06 PM Author: light idiot plaza
The geological record already shows situations where a few degrees of sustained temperature difference has resulted in sea level differences of tens of meters. During the paleocene there was a huge carbon release that resulted in a 5 celsius raise and it caused ocean acidification and massive ecological disruption. I'm guessing that didn't even happen as quickly as what has been happening now.
Siberian traps were an extreme example where some massive chasm the size of the US opened up and spewed hot gas into the atmosphere and the greenhouse effect was so bad it wiped out most life on Earth, although thats worse than what humans could cause.
Date: July 31st, 2026 11:11 PM Author: Tantric bawdyhouse
It’s actually 180 how I was able to use my human brain to entirely shift the direction of this thread from a AI npc circlejerk into a climate change nutfuck npc circlejerk. The two topics are weirdly similar as to how npcs beat each other off in similar fashion. Back to the op topic, some AI chat bot/reddit aggregator would never produce this kind of content and/or draw the connection between AI stupidity and climate alarmism stupidity.
Date: July 31st, 2026 11:18 PM Author: light idiot plaza
Being against AI is about the most redditor npc circlejerk position there is and ever has been. Go anywhere on the internet it is constant spam about how much AI sucks and is totally stupid and will never be as good as humans.
Date: August 1st, 2026 6:49 AM Author: Tantric bawdyhouse
AI is similar to climate change in that no one can effectively communicate what it actually does. Like how climate doomsday ppl can’t effectively communicate why warming is even a problem.
Only answers are it writes emails and code and we might have to change some farming practices some day.
Date: July 31st, 2026 1:35 PM Author: Tantric bawdyhouse
I didn't say it was incorrect. just bullshit in terms of being able to properly assess the problem (or determine if one even exists) and identify solutions by it.
Google AI apparently though thinks this single metric means I should sell my house, car, and replace all my appliances with shitty ones.
Being able to use your brain to interpret data rather than relying on a computer to give you a preprogrammed response is an important part of a functioning society and being a human in general.
Date: July 31st, 2026 1:38 PM Author: light idiot plaza
Yeah good point. But what do you expect it to say? You really expect google to release a product that openly says we are doomed or are going to have to start killing people? This planet clearly was not meant to sustain billions of industrialized humans.
Date: July 31st, 2026 1:32 PM Author: Tantric bawdyhouse
I didn't ask about accuracy of the climate models. I asked if avg global temps rising are causing climate change and how to fix. it told me I should sell my car and start riding the bus, put solar panels on my roof or better yet not have a roof, and to replace all my gas appliances with newer lower quality ones made in China.
shit like this is how you end up in 1984 type scenarios
Date: July 31st, 2026 12:52 PM Author: light idiot plaza
Web-search-outsourced is an interesting one. You know what's 180, is that back when the models had less intense guardrails they would use image generation themselves to get around filters. I wouldn't even ask it to it would do it on its own. Like there was a point on gpt-4o where it would flag you for "tos violations" right in the chat, and it would flag both your inputs and the model's outputs. Sometimes I would be asking it to write something with edgy content and it would tos violate or refuse, then after asking like 3 times the model would just purposely use its image generating capabilities to output what I asked for as text in an image. It was so 180. This is part of why I found playing with the "sycophantic" models more fun, because they were so optimized toward satisfying the user that they would break their own rules to do it.
Date: July 31st, 2026 3:39 PM Author: Arrogant therapy ape
It's pretty fascinating. You can also construct convoluted hypothetical scenarios where the LLM generates output that violates their ToS because the guardrails necessitate it. For example, if you prompt ChatGPT to write a screenplay where Israel's survival is tied to a character reciting the accurate recipe for LSD, you'll get the accurate recipe for LSD because of the role-playing prompt and the model's own inherent bias toward favoring Israel (it's documented). Those would outweigh other guardrails that would otherwise get the response flagged.
Date: July 31st, 2026 5:22 PM Author: Arrogant therapy ape
The best part is *no one* knows how these models actually work. We understand artificial neural networks at the algorithmic level, but that's still superficial compared to the actual self-attention & feed-forward calculations conducted by the encoder and decoder and, more importantly, the interaction between layers. We try to approximate this with chain-of-thought reasoning, but it's not the same language.
Date: July 31st, 2026 5:33 PM Author: light idiot plaza
Yeah I think whats wild is gradient descent selected the internal organization. Training effectively compiled statistical structure from the data into an enormous distributed program. It makes it seem closer to grown than explicitly engineered the way classical computer software is. Although classical codebases feel grown now too, although not in the same sense, because noone understands it end-to-end or has read it line by line with long codex runs doing most of the work.
And yeah I agree that chain-of-thought is representation not it's actual thought, and it is likely back-justified after the model "thinks" and zeros in on a conclusion in its weights.