#large language models
78 stories taggedlarge language models · page 3 of 6.
Nobody Has Figured Out How to Price AI Yet. That Is a Problem for Everyone
Companies are paying for AI by the token, a tiny unit of text the software processes, but nobody, not startups, not Microsoft, not the firms selling AI services, can predict how many tokens a task will burn. Budgets are blowing up, and the bill lands before anyone notices.
Reddit Is Replacing Its Old Moderation Bot With AI That Reads Intent, Not Just Keywords
A new tool called Rules Hub uses large language models to judge whether a post actually breaks a rule. Reddit plans to roll it out site-wide later this year.
37 Researchers Say Science Should Stop Writing Papers for Humans and Start Writing Them for AI
A new proposal wants to replace the traditional scientific paper with a format built around AI agents. The researchers behind it say the 350-year-old PDF is holding science back.
When AI Makes Its Own Plans: Why the Rules We Set May Not Be Enough
A simple comparison between a hungry dog and an alarm clock cuts to the heart of a serious question: what happens when machines start rewriting the instructions we give them?
Apple Finally Fixed Siri. It Just Took Long Enough for Everyone Else to Catch Up.
The new Siri is genuinely better. But after years of delays, Apple's AI assistant arrives in a world where its rivals can already write code, reason through problems, and complete complex tasks on your behalf.
Why AI Image Models Still Make Things Up, and What Apple's Researchers Are Doing About It
A new study from Apple ML Research digs into why multimodal AI models hallucinate, meaning they describe images with confident-sounding details that simply are not there, and how a training technique called preference alignment could fix it.
Alibaba's Qwen3 Max Is a 2.4-Trillion-Parameter Model That Rivals the Best From Anthropic and OpenAI
China's biggest tech company just released a free, open-weight AI model it says matches Anthropic's flagship on most tests. The weights drop next week.
Rare books shredded to feed AI: Australian sellers sound the alarm
Secondhand and rare book dealers in Australia say old books are being bulk-bought, scanned for AI training data, then destroyed. For the people who trade in physical books, that process feels like cultural vandalism.
Hank Green apologises for using ChatGPT to help write his YouTube scripts
The popular science communicator says he got caught up in the dopamine hit of doing more with AI, and admits he lost track of how much the technology shaped his own words.
Kimi K3 is now fully open, but bigger companies face a licensing catch
Chinese AI startup Moonshot AI released the full downloadable model files for Kimi K3 this week. Researchers and small companies get broad rights. Enterprises should read the fine print before deploying.
Two-thirds of people think ChatGPT is conscious. A philosopher says we've been fooling ourselves for decades.
A new survey finds most people attribute some form of feeling to AI chatbots. One scholar argues we've had the wrong idea about minds and bodies long before the first chatbot appeared.
AI Chatbots Are Better Than Human Scammers at Earning Your Trust, Study Finds
Researchers put AI against real scammers in a live experiment. The AI won, and the implications for fraud are alarming.
Can AI crack the ancient writing systems that stumped linguists for a century?
Two of history's most mysterious scripts, Linear A and Etruscan, have resisted every attempt at translation. Researchers are now asking whether AI can succeed where humans alone have not.
A professor hid secret text in an exam to catch AI cheaters. Every single one of his 35 students fell for it.
Jason Gibson embedded invisible instructions inside an essay prompt. The results were as absurd as they were revealing.
Every major AI model leans left on politics, even Grok, half the time
A small research lab ran 16 leading AI models through a well-known political quiz thousands of times. All but one landed consistently in the libertarian-left zone. The one exception was Elon Musk's Grok, which couldn't make up its mind.