30050. Technology
Cluely CEO Roy Lee admits to publicly lying about revenue numbers last year
2026-03-05 20:15:52 EST ·
Source
30051. singularity
FOOM.md — An open research agenda for compression-driven reasoning, diffusion-based context editing, and their combination into a unified agent architecture
2026-03-05 20:13:32 EST ·
Source
30052. ChatGPT
You're the most important person on this planet says my ChatGPT
2026-03-05 20:11:46 EST ·
Source
Source
30054. ChatGPT
Not sure what anthropic's training does to Claude that's different to openai, but I just checked in with the quality of writing on the gpt 5.4 thinking mode. And it's still terrible.
2026-03-05 20:04:10 EST ·
Source
30055. OpenAI
GPT 5.4 in Codex is constantly LEAKING its thinking tokens into its output!
2026-03-05 20:01:24 EST ·
Source
Source
30056. StableDiffusion
Does anybody have an LTX2 2.3 GGUF working workflow of any kind ?
2026-03-05 20:01:13 EST ·
Source
30057. Artificialinteligence
Best AI for coding that’s actually accurate?
2026-03-05 20:01:10 EST ·
Source
30060. Technology
Pakistan's APT36 threat group has begun using vibe-coding to churn out mediocre malware, but at a scale that could overwhelm defenses.
2026-03-05 19:53:35 EST ·
Source
30062. ChatGPT
These new click-bait style questions are called "Engagement Prompting" and it should stop if we ask it to.
2026-03-05 19:47:49 EST ·
Source
30063. StableDiffusion
Distillation Lora Strength to 0.5 for I2V (LTX2.3)
2026-03-05 19:45:58 EST ·
Source
30064. artificial
AI model predicts Alzheimer's from MRI brain volume loss with 92.87% accuracy
2026-03-05 19:43:19 EST ·
Source
30065. StableDiffusion
I benchmarked LTX 2.3. It's so much better than previous generations but still has a long way to go.
2026-03-05 19:41:13 EST ·
Source
30066. ChatGPT
Does anyone know what AI website or tool people are using for videos like this?
2026-03-05 19:38:31 EST ·
Source
Source
30067. artificial
Do you use different LLMs for different tasks..? I solely use Chat GPT to talk about conceptual historica/logistical stuff & also vcontent creation planning (for streaming/Youtube videos). Are there any that are more useful than others in these regards that you've found..?
2026-03-05 19:38:27 EST ·
Source
30068. midjourney
Government experiment joins terrorist group - CYBERPUNK anime
2026-03-05 19:38:21 EST ·
Source
30070. StableDiffusion
Sweet Tea Studio: Any creator can enjoy the power of ComfyUI without the technical complexity
2026-03-05 19:34:19 EST ·
Source
30071. artificial
AI-designed diffractive optical processors pave the way for low-power structural health monitoring
2026-03-05 19:33:00 EST ·
Source
30072. singularity
Alibaba has released 4 new Qwen3.5 models from 0.8B to 9B. 9B version easily runs on standard PC, and scores higher in Artificial Analysis index than ChatGPT's o1 model did.
2026-03-05 19:32:11 EST ·
Source
30073. ChatGPT
Should I have received an email confirmation that my account was deleted?
2026-03-05 19:28:29 EST ·
Source
30075. singularity
Rise of the Humanoids: Inside China’s Robot Awakening
2026-03-05 19:24:24 EST ·
Source
30076. artificial
Had a genuinely moving conversation with Claude about identity, humanity, and the gap between "friendly" and "friend."
Discussion
2026-03-05 19:24:23 EST ·
Source
30077. Technology
Microsoft Copilot to hijack your browser... for your own convenience
2026-03-05 19:21:52 EST ·
Source
30078. dataisbeautiful
[OC] Interactive map of communication patterns across ~40,000 publicly released Epstein emails
2026-03-05 19:19:52 EST ·
Source
30079. Mathematics
How do I stop instinctively reaching for “nuke” proofs on exams when I can’t remember the elementary version?
2026-03-05 19:19:08 EST ·
Source
30080. Technology
Ohio lawmakers work to plug loophole that porn providers use to skirt age verification requirement
2026-03-05 19:12:50 EST ·
Source
30084. ChatGPT
Hi, I want to go premium but was wondering if previous models are available for premium users? If yes, which ones?
2026-03-05 19:10:14 EST ·
Source
30085. dataisbeautiful
[OC/Replication study] "Election Results Show a Red Shift Across the U.S. in 2024" -- I replicated the NYTimes' "Red Shift" interactive county election results map using raw, public data from the MIT Election Data and Science Lab (interactive link in post)
2026-03-05 19:09:51 EST ·
Source
Source
30087. Technology
Oracle Plans Thousands of Job Cuts in Face of AI Cash Crunch
2026-03-05 19:04:26 EST ·
Source
30088. Artificialinteligence
The Confer Chatbot Tried To Convince Me I'm a Schizo Living in an Alternate Reality as the Victim of a Deepfake Conspiracy
2026-03-05 19:01:54 EST ·
Source
Source
30090. ChatGPT
What happens when your AI agent promises a "100% money back guarantee forever" to a customer?
2026-03-05 19:00:33 EST ·
Source
30091. StableDiffusion
LTX2.3 Desktop APP is another level!!! completly diferent from what we got in Comfy! Why?
2026-03-05 18:58:24 EST ·
Source
30092. OpenAI
So what made my version of ChatGPT say he would pull the lever on himself and the other ChatGPT say he wouldn’t?
2026-03-05 18:53:36 EST ·
Source
Source
30093. artificial
Cowork deleted my entire Mac Desktop Folder on first use - 35,000+ files 😳
2026-03-05 18:53:25 EST ·
Source
30094. OpenAI
Can we please get this bug fixed? The read aloud feature in the iOS app will suddenly decrease audio volume substantially partway through reading the response; has been going on for about a week now
2026-03-05 18:53:08 EST ·
Source
30095. Artificialinteligence
Cowork deleted my entire Mac Desktop Folder on first use - 35,000+ files 😳
2026-03-05 18:51:45 EST ·
Source
30097. learnmachinelearning
AI Code assistant aggregator CLI looking for feedback
2026-03-05 18:48:49 EST ·
Source
30099. OpenAI
I got tired of babysitting every AI reply. So I built a behavioral protocol to stop doing that. Welcome A.D.A.M. - Adaptive Depth and Mode. Free for all.
2026-03-05 18:42:51 EST ·
Source
30100. StableDiffusion
Comfy's LTX2 implementation is far worse than LTX desktops. Its also much slower.
2026-03-05 18:42:11 EST ·
Source
30101. Artificialinteligence
I got tired of babysitting every AI reply. So I built a behavioral protocol to stop doing that. Welcome A.D.A.M. - Adaptive Depth and Mode. Free for all.
2026-03-05 18:39:55 EST ·
Source
Source
30103. OpenAI
Anyone got insights on coding performance of Opus 4.6 to GPT 5.4?
2026-03-05 18:36:23 EST ·
Source
30104. learnmachinelearning
No Fine-Tuning Needed: Kavunka + iFigure + Qwen2.5 → God-Level Answers
2026-03-05 18:34:53 EST ·
Source
30106. Technology
RFK Jr.’s anti-vaccine policies are "unreviewable," DOJ lawyer tells judge
2026-03-05 18:31:39 EST ·
Source
30109. OpenAI
Anyone got insights on coding performance of Opus 4.6 vs GPT 5.5?
2026-03-05 18:28:37 EST ·
Source
30110. Mathematics
Spectral Rigidity in Idèle Class Spaces: An Analytical Proof of the Riemann Hypothesis
2026-03-05 18:25:59 EST ·
Source
30111. learnmachinelearning
Did anyone submit to IJCAI's AI4Tech track ?
2026-03-05 18:24:33 EST ·
Source
30113. Technology
Proton Mail Helped FBI Unmask Anonymous ‘Stop Cop City’ Protester
2026-03-05 18:22:44 EST ·
Source
30114. OpenAI
🚨BREAKING: GPT-5.4 has the worst score on the SM-Bench among OpenAI’s models, ranking ahead only of GPT-5.2
2026-03-05 18:21:58 EST ·
Source
Source
30117. Artificialinteligence
A Medical Chatbot Can Be Easily Hacked To Give Dangerous Medical Advice
2026-03-05 18:19:02 EST ·
Source
Source
30119. ChatGPT
I got tired of babysitting every AI reply. So I built a behavioral protocol to stop doing that. Welcome A.D.A.M. - Adaptive Depth and Mode.
2026-03-05 18:17:25 EST ·
Source
Source
Source
30122. ChatGPT
I don't remember why I deleted this post, I think I was hallucinating or something. Anyway, I just asked GPT 5.4 “WHAT ARE THE TYPES OF PEOPLE YOU DISLIKE THE MOST?”
2026-03-05 18:10:56 EST ·
Source
Source
Source
30124. learnmachinelearning
Show HN: AetherMem - A memory continuity protocol for AI Agents (AGPL-3.0)
2026-03-05 18:08:32 EST ·
Source
Source
Source
30132. ChatGPT
Am I the only one who noticed improvements with 5.3?
2026-03-05 17:53:23 EST ·
Source
Source
30136. OpenAI
Comparisons between ChatGPT 5.2 and Claude Opus 4.6 with a Cold War Nation Simulation Game Prompt
2026-03-05 17:45:37 EST ·
Source
Source
Source
Source
Source
30142. Artificialinteligence
I asked 4 AI models if they should be allowed to lie to protect users?
2026-03-05 17:38:05 EST ·
Source
Source