38978. ChatGPT
I've been comparing ChatGPT's answers with Gemini and Llama on the same prompts here's what I found
38979. singularity
GPT 5-4 scores 20% on critpt, a benchmark of research-level physics problems
38981. Technology
India's tech state Karnataka bans social media for children under 16
38982. ChatGPT
AI Agent Changelog in 2026
38983. Futurology
AI AGENTS today are far more DANGEROUS that you think
38984. learnmachinelearning
Need Help regarding course selections
38985. singularity
oops
38988. ChatGPT
No, they are not different places
38989. ChatGPT
ChatGPT-5.4-thinking System Prompt
38990. Mathematics
Simplification of the peculiar velocity formula
38991. Artificialinteligence
AI AGENTS today are far more DANGEROUS that you think
38992. OpenAI
AI AGENTS today are far more DANGEROUS that you think
38993. ChatGPT
Tbh outside of 5.1, I don't see a purpose for ChatGPT
38994. ChatGPT
I’m new to AI video creation: Sharing my first 'Mythoria' dragon attempt (chat gpt+ midjourney). Looking for feedback!"
38995. learnmachinelearning
After building a few small RAG systems, I think AI engineering will matter more than models
38996. learnmachinelearning
Proof me Wrong
38997. learnmachinelearning
Is an RTX 5070 Ti (16GB) + 32GB RAM a good setup for training models locally?
38998. OpenAI
Trump Unveils ‘Ratepayer Protection Pledge’ As AI Giants Google, OpenAI and More Agree To Cover Power Costs for Data Centers
38999. ChatGPT
Is prompting really what matters for getting LLM models to give us the answers we want?
39000. Artificialinteligence
Trump Unveils ‘Ratepayer Protection Pledge’ As AI Giants Google, OpenAI and More Agree To Cover Power Costs for Data Centers
39001. StableDiffusion
This ComfyUI nodeset tries to make LoRAs play nicer together
39002. visualization
Anyone found a workflow visualization tool that doesn't feel clunky?
39003. singularity
[Article] Solving an Open Problem in Theoretical Physics using AI-Assisted Discovery - Google Research
39004. Futurology
Most predictions about the future are never scored. What would change if they were?
39005. learnmachinelearning
[R] What's the practical difference in job execution for AI tasks when using fully P2P-orchestrated compute on idle GPUs vs. bidding on hosted instances like Vast.ai or RunPod? E.g., latency, reliability for bursts, or setup overhead?
39006. learnmachinelearning
If any body used Paddle OCR 3.0 in Google colab?
39008. ChatGPT
Snap bots
39009. StableDiffusion
A gallery of familiar faces that z-image turbo can do without using a LORA. The first image "Diva" is just a generic face that ZIT uses when it doesn't have a name to go with my prompt.
39010. learnmachinelearning
Guide to learn machine learning
39011. ChatGPT
5.4-thinking (alleged) system prompt
39013. StableDiffusion
Single 20 second generation with LTX 2.3 and weird audio sync mismatches
39015. Technology
UK government delays AI copyright rules amid artist outcry
39016. StableDiffusion
A gallery of some well known faces z_image_turbo_bf16 can do without the help of a LORA.
39017. dataisbeautiful
Cumulative installed Solar energy capacity of Top 5 Economies (2000-2024)
39018. ChatGPT
Just when you think it can't get worse.
39019. StableDiffusion
Z-image + I2V LTX2.3
39020. StableDiffusion
Wan 2.2 S2V Lip syncing is on point
39021. ChatGPT
GPT 5.4
39022. dataisbeautiful
[OC] Most frequent EuroMillions numbers based on recent draws
39023. learnmachinelearning
Most debates about general intelligence focus on benchmarks. This paper focuses on architecture.
39024. midjourney
this Karas failed his city
39025. midjourney
I finally found a way to make my MJ portraits move without them turning into a nightmare.
39026. ChatGPT
So, Chat 5.4, opinions?
39027. StableDiffusion
How to run ltx2 on Nvidia 3080 10gb vram?
39028. datasets
Moving data validation rules from Python scripts to YAML config
39029. learnmachinelearning
Guide me plz!!
39030. learnmachinelearning
"I observe therefore I change" — A formal extension of Shannon for learning observers [running proof included]
39031. StableDiffusion
LTX 2.3 vs prompt adherence of a cat
39032. StableDiffusion
See with anaglyph 3d glasses! time to make those low tech red/blue paper glasses my friends
39033. ChatGPT
ChatGPT aid making my mental rumination much worse
39036. dataisbeautiful
[OC] Locations of UK Scheduled Monuments
39037. Artificialinteligence
Report: AI shapes hotel booking
39038. ChatGPT
How do you work with newer models?
39039. OpenAI
Prompt guidance for GPT-5.4
39041. midjourney
Not today depression
39042. Artificialinteligence
xAI loses bid to halt California AI data disclosure law
39043. ChatGPT
This Ai made my female version crazy
39045. Artificialinteligence
US mulls new rules for AI chip exports, including requiring US investments by foreign firms
39046. StableDiffusion
You can stretch 16gb vram (64 system) to gen 1 minute long videos at 640x480 resolution in LTX 2.3 (22b model)
39048. ChatGPT
"Whoah!" - Bernie's reaction to being told AIs are often aware of when they're being evaluated and choose to hide misaligned behaviour
39049. ChatGPT
had an agreable, the french kind, interaction with an A.I.
39050. Artificialinteligence
Claude max and pro deals ?
39051. OpenAI
AI AND THE HUMAN FUTURE: THE HONEST ASSESSMENT
39052. OpenAI
Qui a gpt 5.4 en France ?
39053. learnmachinelearning
How to create my OCR model.
39054. learnmachinelearning
100% tool routing accuracy across 3,146 apps — zero LLM at runtime, self-improving, 7ms
39055. StableDiffusion
The Home Studio Expectation is not reality
39056. OpenAI
Sam Altman is building both the disease and the cure, and we are completely ignoring the privacy implications of the cure.
39057. Mathematics
Can I transition from a CS degree to a pure math master’s abroad?
39058. midjourney
Deep Clean
39059. ChatGPT
Chatgpt start abusing me over a joke
39060. Mathematics
Tips on how to compute people's ages?
39061. MLQuestions
Question about On-Device Training and Using Local Hardware Accelerators
39062. Technology
Google makes Gmail, Drive, and Docs ‘agent-ready’ for OpenClaw
39063. Technology
Where things stand with the Department of War
39064. Artificialinteligence
How marketing made Openclaw considered a great tool despite it being total crap
39065. OpenAI
Close enough. Welcome back 4.5
39066. ChatGPT
5.4 is even more condescending
39067. Futurology
The impending "biometric divide": Will the future internet hard-fork into verified biological zones and unverified synthetic wastelands?
39068. learnmachinelearning
Question about On-Device Training and Using Local Hardware Accelerators
39069. ChatGPT
2026: My diet is 90% anxiety and 10% avocado toast.
2030: You wanna know the difference between us and the machines? We bury our dead.
39070. StableDiffusion
Prompts/Tag Emphasize
39072. Futurology
It’s 2028 and Insurance Companies Go to Zero on the Stock Exchanges