Su X l’amministratore delegato di #OpenAI, Sam Altman, ha scritto: “Addio #Gpt-4, hai dato il via a una rivoluzione. Conserveremo con orgoglio i tuoi parametri su uno speciale disco rigido per poterli consegnare agli storici in futuro”.
Il suo addestramento è costato oltre 100 milioni di dollari.
#GPT
In the middle of a small project I was doing with GPT a few weeks ago (failing to convert sheet music to MIDI), I had a good chat with it about ethics, the singularity, the stealing of the libraries and works of humanity, and right-wing bias training of AI models.
It was quite a good chat, so I decided to post it.
At the moment, I am quite optimistic about AI's future with humanity, but I am sure some billionaire owned corporation will come along and fuck that up somehow.
#AI #GPT #Ethics #Codex #Theft #AITraining #Singularity #Books #Music #Copyright #Grok #BillyBragg
The Hundred-Page Language Models Book by Andriy Burkov is on sale on Leanpub! Its suggested price is $50.00; get it for $20.00 with this coupon: https://leanpub.com/sh/zdmLbhYC #Ai #Gpt #NeuralNetworks #DeepLearning #DataScience #ComputerScience #Linguistics
Behold, the glorious #AI Frankenstein: 🤖📺 GPT-FoxNewsDad! For those in #tech and academia who need life #advice from a #virtual #dadbot that probably thinks "illegal content" is anything not reported by Fox News. Just what we all needed, a #digital #dad to remind us to mow the lawn and question our career choices! 🙄
https://chatgpt.com/g/g-6814740bd1cc81919d28a7c4b8db055d-gpt-foxnewsdad #Frankenstein #GPT-FoxNewsDad #HackerNews #ngated
GPT решает профильную математику (ЕГЭ)
Мне давно хотелось проверить текущий уровень языковых моделей от Open ai на целом наборе математических задач, только на уровне ЕГЭ. Я уже имею достаточно большой опыт работы с GPT-o3-mini high, даже та модель справлялась с задачами олимпиадного уровня, о чём будет дальнейшая статья (надеюсь). Но сейчас про ЕГЭ. Как многие знают, профильный ЕГЭ по математике делится на 2 части: 1 с кратким ответом и 2-ая, где необходимо полное и обоснованное решение.
All Books from The Medical Futurist https://leanpub.com/b/tmf-books by Dr. Bertalan Mesko is the featured bundle of ebooks 📚 on the Leanpub homepage! https://leanpub.com #Medicine #Leadership #ComputerProgramming #Ai #DigitalTransformation #Gpt #Investment #books #ebooks
Convert Plain English To Commands Using GPT-3 Powered Shell Genie https://www.linuxuprising.com/2023/02/convert-plain-english-to-commands-using.html #linux #shell #gpt #ai
Stop paying for idiots' #clone via their #GPT, #GPTAgent in #ChatGPT, or any LLM, to obtain their expertise 24/7! Such is dipshittery, as u can get freely by uploading their online data/sources in ur FREE account! This is confirmed even by ChatGPT itself!
Wake up, u planet of dumbshit humans!
Sometimes I see anti-AI posts that legit baffle me. Like, I can definitely understand if you're anti-AI for philosophical reasons. I don't agree, but I understand your worldview. What I don't get is just saying it's bad at everything. Legit seems like these people haven't used an AI since GPT-2 and DALL-E.
GPT Group ( #GPT ) has released " 2025 Annual General Meeting Chairman and CEO Addresses " on Thu 01 May at 09:33 AEST #tax #Australia #Strategy #live #ASX
https://grafa.com/asset/gpt-group-2138-gpt.asx?utm_source=asxmktsensitive&utm_medium=mastodon&utm_campaign=gpt.asx
Generative AI in a Nutshell by Henrik Kniberg is on sale on Leanpub! Its suggested price is $39.00; get it for $20.30 with this coupon: https://leanpub.com/sh/kXGX0xlJ #DigitalTransformation #Gpt #Ai
The end of an AI that shocked the world: OpenAI retires GPT-4 https://arstechni.ca/da94Z #largelanguagemodels #machinelearning #AIassistants #samaltman #ChatGPT #chatgtp #Biz&IT #openai #GPT-4 #AI
Four weeks ago we’ve released the first public beta of DEVONthink 4. Today, we’re following up with public beta 2 which brings support for new AI models, new scripts, interface enhancements, and more. #devonthink #devonthink4 #chatgpt #gpt-4.1 #gemini https://www.devontechnologies.com/blog/20250430-devonthink-40-public-beta-2
AI-агенты в реальном мире: почему они не работают и как это исправить
AI-агенты часто не оправдывают ожиданий при внедрении в реальные продукты. Разбираемся, почему сложно оценивать их эффективность, какие ограничения у статических бенчмарков и почему фокус на надёжности важнее, чем на возможностях систем
https://habr.com/ru/articles/905590/
#aiагент #llmмодели #llm #инженерия_надежности #бенчмарки #галлюцинации #оценка #claude #gpt #парадокс_джевонса
[Перевод] Как я использовал ИИ для создания рабочего эксплойта для CVE-2025-32433 до появления публичных PoC
Может ли ИИ просто написать эксплойт за меня? GPT-4 не только понял описание CVE, но и выяснил, какой коммит ввел исправление, сравнил его со старым кодом, нашел разницу, нашел уязвимость и даже написал доказательство концепции. Давайте рассмотрим, как я к этому пришел.
https://habr.com/ru/articles/905552/
#gpt #информационная_безопасность #cve #ai #искусственный_интеллект #cursor #pentest #penetration_testing #poc
Minha opinião é que 50 anos é muito tempo para fazer previsões e, na verdade, é tanto tempo que eu veria como uma crítica às IAs generativas que estariam, na opinião dele, tão longe de serem boas que talvez nunca cheguem a esse ponto.
Почему до сих пор ни один ИИ не может написать даже простой проект сам?
Почему даже Cursor AI не может написать простой проект по моему промпту? Автор: A.Ankalaev Cистемный администратор с опытом более 20 лет
An Acclaimed Book About Digital Manipulation Was Actually Written by #AI - https://www.wired.com/story/an-acclaimed-book-about-digital-manipulation-was-actually-written-by-ai/ "I didn’t ask the machine to write for me, but instead it generated ideas and then I used #GPT and #Claude to critique them, to give me perspectives on what I had written. Everything written in the book is mine."
Client Info
Server: https://mastodon.social
Version: 2025.04
Repository: https://github.com/cyevgeniy/lmst