Software engineer, tech blogger, community lead, AI adoption consultant, teaching devs how to work with AI. Belarusian originally 🤍❤️🤍itbeard.com PolandJoined April 2023
Claude in Excel is now available on Pro plans.
Claude now accepts multiple files via drag and drop, avoids overwriting your existing cells, and handles longer sessions with auto compaction.
Get started: claude.com/claude-in-excel
«Цёмны Лёс» аднаўляе працу! І гэта, мабыць, самы гарачы і насычаны выпуск🔥
👉 Глядзім і слухаем тут: youtu.be/_2k5HiDKbLU
Буду бясконца ўдзячны за вашыя рэтвіты і лайкі! Дапамажыце распаўсюдзіць беларускае слова, каб нас пачула як мага больш людзей 🔄❤️
How is this even possible? The whole X and the web were talking about the “open-sourced” MiniMax M2.1, but model does not actually exist as open source. What kind of creepy marketing is this? Even the official documentation for the model on the vLLM website has a 404 link to HF..
@AGulev Of course, it needs to be used depending on the situation. However, when working with a large amount of data, saving just 2 tokens out of 38 (approximately 2.5%) can cost a company thousands of dollars.
The first preview of "Group chats" in the ChatGPT web app
- the top navigation bar now has a "Start a group chat" button which allows you to generate a link and share it with others to join this group chat shown in the new sidebar section "Group chats"
- anyone can join your group chat with this link, and they'll be able to see the previous messages in this group chat
- group chats custom instructions are separate from custom instructions for your personal ChatGPT, and you can let ChatGPT respond automatically or only when mentioned by someone in the group chat, but your personal ChatGPT memory is never used in group chats
- the web app assets also mention options to add reactions, report or reply to a specific message, an indicator when someone is typing, file uploads, image creation and web search
Are u already know about way more efficient alternative to JSON for working with models? Now you are - say warm welcome TOON (Token-Oriented Object Notation): github.com/toon-format/to….
It gives pretty cool results while tokenization 👍
Google Gemini 3 will be announced on Wednesday, but the release date will be November 12, according to this leaked release plan. Of course this could be changed at any time, especially after this leak.
Gemini 3.0 Pro, November 12th
4 Followers 231 FollowingObsessed with innovation? So are we! 🔥 Elektra Home is your ultimate destination for everything tech: from the newest iPhones & Macs to powerful PCs, smart hom
982 Followers 122 FollowingAI Agent Developer. I build Complex Agents and teach people how to build AI agents from scratch with a focus on IDE-agnostic architecture.
4K Followers 268 FollowingSoftware developer. Ex VK, Telegram. Currently working on two Fediverse projects. Deeply disappointed with the IT industry. AI hater.
3K Followers 2K FollowingDOBOT is a global leading provider of all-perceptive intelligent robotics solutions, focusing on the development of all-perceptive intelligent robotics.
2K Followers 162 FollowingProduct guy, systems thinker & educator. Love building stuff. Founder @appintheair. Now building AI agents for b2b sales at https://t.co/St0me2s4Pv
11 Followers 13 FollowingBuilding sharp infrastructure for databases, code, and AI. Creator of pg_sorted_heap, depyo, cogni-ml, and Crystal Lang V2.
GitHub: https:/github.com/skuznetsov