michabbb

#devop, #php, #laravel, #mysql, #api expert, #ebay #developer, #openai, #ai, #llm,. #livewire, #alpinejs, searchable

I built https://burner-note.link🔒⏳🔥

Public Key

npub1z20c8zvvwqydxdthrlngrm8e08nhv7ketrz49lu9m6tzhghhwklql84yd9

NIP-05 Address

michabbb@social-vivaldi-net.mostr.pub

Profile Code

nprofile1qqsp98ur3xx8qzxnx4m3le5panuhnemk0tv4332jl7zaa93t5tmht0spz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dslxf7sd

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1z20c8zvvwqydxdthrlngrm8e08nhv7ketrz49lu9m6tzhghhwklql84yd9

Show more details

Last Notes

2026-01-14 15:43:10 UTC

📦 Efficient file syncing for massive datasets? Forget classic rsync!
⚡ #BLAKE3 hashes instead of timestamps – hash 70GB in just ~20 seconds thanks to multithreading across all CPU cores
#DevOps #Linux #SysAdmin #OpenSource #FileSync
📊 Algorithm comparison for 70GB:
- BLAKE3: ~20 seconds
- MD5: ~1.5 minutes
- SHA256: ~4.5 minutes
🔍 The trick: Create a hash baseline, compare with `comm -13`, then use `rsync --files-from` to transfer only actually changed files
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/894/215/488/924/512/original/d09c0f9c858aa345.png

2025-10-05 08:08:05 UTC

- reply

🔄 Built-in retry mechanism to handle database deadlocks automatically with configurable attempts
🎛️ Multiple database connection support for complex multi-database applications
⚙️ Requires #PHP 8.1 or higher and #Laravel 10.x or higher for installation
🔌 Simple installation via #Composer with automatic service provider registration
📝 Four implementation methods: trait with runInTransaction, middleware for controllers, helper functions, and attributes

2025-10-05 08:08:04 UTC

📦 #Laravel Auto Transaction package eliminates manual database transaction management #PHP #database #opensource #webdev
🛡️ Middleware support for automatic transactions across entire routes with custom connection options
🔧 Trait integration using HandlesTransactions for easy implementation in services and repositories
🎨 Helper functions including transactional() and auto_transaction() wrappers for simple operations
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/320/532/787/044/575/original/972f2237e7fa4742.png

2025-10-03 18:54:17 UTC

- reply

🖥️ Communicates with Windows using WinBoat Guest Server to retrieve necessary data and manage applications running in the containerized environment
📋 System requirements: At least 4GB RAM, 2 CPU threads minimum, 32GB free storage space in /var directory, KVM virtualization enabled in BIOS/UEFI
🐳 #Docker installation required for containerization, Docker Compose v2 for compatibility, user must be added to docker group for proper permissions

2025-10-03 18:54:17 UTC

- reply

🚫 Docker Desktop unsupported, #Podman unsupported, distributions emulating Docker through Podman socket unsupported, rootless containerization solutions currently unsupported
🛠️ Built with #NodeJS and #Go, open source and accepts contributions for bug fixes, feature improvements, documentation updates and general feedback
🌟 Inspired by similar projects including #WinApps, #Cassowary, and dockur/windows which is also used within WinBoat architecture

2025-10-03 18:54:17 UTC

- reply

🖧 #FreeRDP version 3.x.x with sound support required for remote desktop connection, iptables and iptable_nat kernel modules must be loaded
💾 Available download formats: #AppImage for portable usage across distributions, Unpacked raw files, .deb packages for #Debian based distributions, .rpm packages for #Fedora based distributions
⚠️ Currently in beta phase - expect occasional bugs and issues, requires some troubleshooting comfort level from users testing the application

2025-10-03 18:54:16 UTC

- reply

✨ Smartcard passthrough support, comprehensive resource monitoring dashboard, and many more features being added regularly to enhance functionality
🔧 Technical approach: #Electron application running Windows as virtual machine inside #Docker container, uses #FreeRDP together with Windows RemoteApp protocol for native window compositing and seamless application integration

2025-10-02 11:15:21 UTC

Unsloth now has a #Docker image! #ai
#Train #LLM locally with no setup: just run the image and go. Includes every pre-made Unsloth notebook.
Solves dependency or environment issues.
Guide: https://docs.unsloth.ai/new/how-to-train-llms-with-unsloth-and-docker
https://social-cdn.vivaldi.net/system/media_attachments/files/115/304/282/079/458/105/original/74ada35d3377e698.jpeg

2025-10-01 00:45:01 UTC

- reply

🎯 Real-world validation through extended #CCBench testing with human evaluators completing multi-turn tasks in isolated #Docker containers across frontend development, tool building, data analysis, testing & algorithms
🔧 Near parity with #ClaudeSonnet4 (48.6% win rate) while outperforming other #opensource baselines in practical scenarios
⚙️ 15% more token-efficient than #GLM45, finishing tasks with fewer tokens while maintaining higher capability levels

2025-10-01 00:45:00 UTC

🚀 Welcome GLM-4.6 the Latest flagship #opensource #AI #llm with advanced agentic, reasoning & coding capabilities
⚡ Performance improvements over #GLM45 with competitive advantages against #DeepSeekV3 and #ClaudeSonnet4 across 8 public benchmarks covering agents, reasoning & coding
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/296/141/055/584/770/original/fca7bfb1850456d8.png
https://social-cdn.vivaldi.net/system/media_attachments/files/115/296/141/108/679/323/original/e50776ba6ab313bd.png

2025-09-04 19:43:52 UTC

Give Your #AI more Guidance in Writing #PHP code
With 189 Custom #phpstan rules 🚀 🔥
As a #phpstorm User, you should know phpinspectionsea -
if not, shame on you! 😏
#Ai shouldn't need an IDE to find errors or warnings in the code.
Because #jetbrains #MCP became useless to me and I only used 3 tools anyway, I let Ai create 189 custom #phpstan rules based on phpinspectionsea 😏
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/147/737/145/279/657/original/5c082a66b8ccf974.png

2025-08-27 10:01:25 UTC

📋 #Flowforge Transform any #Laravel model into production-ready drag-and-drop #Kanban board 📋
⚡ Handles 100+ cards per column with intelligent pagination and optimistic UI
🎯 Three integration patterns: #Filament Pages, Resources, or standalone #Livewire
🔧 Zero configuration required - works with existing models and database
🎨 Fully customizable with actions, schemas, filters, and themes
🚀 90-second quick start with single artisan command board generation
🧵 👇 #php
https://social-cdn.vivaldi.net/system/media_attachments/files/115/100/147/988/348/764/original/7565db089e93f51e.mp4

2025-08-26 14:45:09 UTC

- reply

🌍 World Knowledge Integration
Benefits from Gemini's world knowledge for semantic understanding beyond aesthetic generation. Enables educational applications and real-world question assistance.
🖼️ Multi-Image Fusion
Understands and merges multiple input images:
Place objects into scenes
Restyle rooms with color schemes or textures
Create photorealistic fused images with single prompts
🔗 Platform Partnerships
https://OpenRouter.ai: First image generation model among 480+ models available

2025-08-26 14:45:09 UTC

- reply

🔧 Key Features
Character Consistency
Maintains appearance of characters or objects across multiple prompts and edits. Enables placing the same character in different environments, showcasing products from multiple angles, or generating consistent brand assets.
✏️ Targeted Transformations
Precise local edits using natural language commands:
Blur backgrounds
Remove objects or stains
Remove people from photos
Alter poses
Add color to black and white photos

2025-08-25 11:43:15 UTC

- reply

🔧 Strict-Mode Function Calling
Beta API now supports strict-mode for function calling, ensuring the output always complies with the defined JSON schema.
🔗 Anthropic API Compatibility
The API is now compatible with Anthropic API, allowing seamless integration with Claude Code.
💰 Pricing Adjustment (Effective Sept. 5, 2025, 16:00 UTC)
Input tokens: $ 0.07/million (cache hit) / $ 0.56/million (cache miss)
Output tokens: $ 1.68/million
The off-peak discount ends simultaneously

2025-08-25 11:43:15 UTC

- reply

Current pricing remains until effective date
📈 Service Resources
Scaled up service resources to better meet API needs. Users are encouraged to utilize the enhanced service.
📚 Documentation
For more details, visit DeepSeek API Docs:
https://api-docs.deepseek.com

2025-08-23 17:54:41 UTC

✨ Say goodbye to messy Laravel configs – embrace type safety! ✨
🚀 #Laravel configs don’t have to be messy. Magic strings? Gone. Type safety? In.
🔑 With #PHP enums + Laravel’s new #[Config] + #[Singleton] attributes, you can map your config values straight into a clean, readonly class.
💡 No more config('aim.api_token') sprinkled everywhere. Instead, inject a typed object and keep your code predictable & elegant.
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/079/360/098/419/936/original/55a8b5eb6604c8c5.png

2025-08-16 00:25:06 UTC

- reply

Required to access usage dashboard
Older versions may consume allocation without notice
Essential for managing pricing plans
🌍 Billing Notes
💸 Tax Information
Prices exclude applicable taxes (VAT, sales tax)
Japanese billing addresses subject to Consumption Tax
Tax amount depends on billing address and Tax ID
📅 Usage Policy
Monthly limits reset at billing cycle start
No rollover of unused requests
Individual subscriptions only (team billing coming soon)

2025-08-16 00:25:05 UTC

- reply

125 spec requests
Pay-per-use overage available
Ideal for regular developers
⭐ Kiro Pro+ - $40/month
450 vibe requests
250 spec requests
Pay-per-use overage available
Enhanced capacity for active users
🔥 Kiro Power - $200/month
2,250 vibe requests
1,250 spec requests
Pay-per-use overage available
Enterprise-level usage
🎯 Key Features Launched Today
📊 Usage Tracking
Live tracking of spec and vibe requests in IDE
Real-time usage dashboard

2025-08-16 00:25:05 UTC

- reply

🔧 Request Types
Vibe Requests: Conversational interactions, coding questions, documentation
Spec Requests: Structured development workflow executions
Clear metering system replacing complex token billing
⚡ Welcome Bonus Details
🎁 14-Day Trial
100 spec requests + 100 vibe requests
Clock starts on first Kiro usage
Available regardless of chosen tier
Unused requests roll over when upgrading
📋 Important Requirements
🔄 IDE Update Required
Update to Kiro IDE v0.2.13 or newer

2025-08-16 00:25:04 UTC

#Kiro introduces new #pricing plans with free tier and premium options 🚀
#ai #coding #aws
🚀 New Pricing Launch
Kiro transitions from preview to paid model with comprehensive tier structure. All existing accounts migrated to Kiro Free plan with billing options now available.
💰 Pricing Tiers
🆓 Kiro Free - $0/month
50 vibe requests monthly
14-day welcome bonus: 100 spec + 100 vibe requests
Perfect for getting started with Kiro
💼 Kiro Pro - $20/month
225 vibe requests
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/035/596/684/881/080/original/6a7e72bf905288a7.png

2025-08-15 14:47:31 UTC

#NGINX introduces native #ACME protocol support for automated #SSL certificate management 🔐 #DevOps
🔧 Built-in ngx_http_acme_module provides directives for requesting, installing & renewing certificates directly from #NGINX configuration
⚡ Eliminates reliance on external tools like #Certbot, reducing manual errors and vulnerabilities
🦀 Leverages #Rust SDK available as dynamic module for both open source and #NGINXPlus enterprise users
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/033/325/663/938/253/original/a06a02bbaac4f2cd.png

2025-08-14 09:30:11 UTC

- reply

- 📹 SRT subtitle creation
- 🔄 Workflow automation

2025-08-14 09:30:10 UTC

🎵 #FFmpeg 8.0 integrates #Whisper for local audio transcription 🎵 #ai
🔧 Direct integration of #OpenAI's #Whisper speech recognition system within #FFmpeg framework
🌐 Completely local processing - no cloud transfer required, ensuring privacy and data security
🗣️ Automatic language detection supporting transcription in over 90 languages worldwide
⚙️ GPU acceleration enabled by default with whisper.cpp library requirement
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/026/415/444/121/028/original/9324fdc61daf0ff8.png

2025-08-14 09:30:10 UTC

- reply

📹 Creates SRT subtitle files for videos and supports real-time live broadcast transcription
🔄 Seamless workflow integration
🔄 allowing automated processing and data transfer to other applications
🔗 Source URL
https://www.heise.de/en/news/FFmpeg-8-0-integrates-Whisper-Local-audio-transcription-without-the-cloud-10522091.html
📊 Key Features Highlighted
- 🔧 Direct FFmpeg integration
- 🌐 Local processing (no cloud)
- 🗣️ 90+ language support
- ⚙️ GPU acceleration

2025-08-13 09:46:40 UTC

- reply

🤖 Context-Aware Agents
Build agents maintaining context across hundreds of tool calls and multi-step workflows with complete #API documentation, tool definitions, and interaction histories without losing coherence.
💰 Pricing Structure
Standard rates: Prompts ≤200K tokens ($3/MTok input, $15/MTok output)
Extended context: Prompts >200K tokens ($6/MTok input, $22.50/MTok output)
Additional savings: Combined with prompt caching and batch processing for 50% cost reduction

2025-08-13 09:46:39 UTC

Claude Sonnet 4 - 1M Context Window Content Structure
🚀 Main Announcement
#Anthropic announces #ClaudeSonnet4 now supports up to 1 million tokens of context - a 5x increase that transforms how developers work with #AI for large-scale projects.
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/115/020/817/971/119/380/original/ee52182916a0f848.png

2025-08-13 09:46:39 UTC

- reply

🔧 Large-Scale Code Analysis
Process entire codebases with 75,000+ lines including source files, tests and documentation for complete system understanding. Claude can identify cross-file dependencies and suggest improvements accounting for complete system design.
📄 Document Synthesis Capabilities
Analyze hundreds of legal contracts, research papers or technical specifications while maintaining full context across documents. Process extensive document sets and analyze relationships

2025-08-11 20:42:23 UTC

- reply

🍒 Cherry-pick commits easily with shift+c to copy and shift+v to paste. Create and manage #worktrees to work on multiple branches simultaneously. Custom command system for extending functionality beyond built-in features.
🎨 Visual commit graph showing relationships and author colors. Built-in support for #Gitflow workflow and branch management. Filter any view to quickly find what you need.

2025-08-09 20:14:41 UTC

do you have the same fun like me using #gpt5? 🙄
#ai #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/115/000/638/267/338/837/original/451407382e5e9ee9.png

2025-08-08 18:30:32 UTC

- reply

@nprofile…5epm phpstorm is the best for php, but they should stop trying AI... they suck super hard at this field.....

2025-08-08 09:04:11 UTC

#cursorcli + #gpt5 = comparing two small markdown files and finding differences, and adding missing parts to the other file
this took ~5 minutes 🥱
holy shit.....
#ai #coding

2025-08-07 10:08:44 UTC

Oh wow... 4 months later, #jetbrians has enough trust to claim: #junie "should be working" now with #WSL2 - I don´t believe it. what a joke!
4 months in a world of #ai is a decade.... poeple are using other stuff, why should someone care now.... again: what a very very bad joke!
#php
https://social-cdn.vivaldi.net/system/media_attachments/files/114/986/930/919/911/876/original/82f8564e7f9c4373.png

2025-08-04 21:01:59 UTC

#Laravel #Livewire 4: The Future of #PHP Components 🚀
🔧 Single-file components with native #JavaScript support - no more script directive needed
⚡ New lightning bolt file naming convention for instant visual recognition
🏗️ Multi-file component option with automatic code-splitting and caching
📁 Opinionated folder structure reduces decision fatigue for developers
🎯 Deep integration with #PHP 8.4 property hooks for validation and computed properties
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/114/972/512/394/607/126/original/3ae87225e7b93a3b.png

2025-07-31 13:12:09 UTC

- reply

@nprofile…v998 that's what many think...

2025-07-31 12:27:38 UTC

Introducing Horizon Alpha, a new stealth #LLM 🌅
currently #FREE 👀
https://openrouter.ai/openrouter/horizon-alpha
#ai #coding #openai ?
https://social-cdn.vivaldi.net/system/media_attachments/files/114/947/840/609/824/159/original/4b4469c91774a3cc.png

2025-07-30 23:04:20 UTC

My #Series Recommendation 📺
👉 STICK ⛳ 🏌️
if you love Owen Wilson, you will love this funny "feel good" series ❤️
https://www.imdb.com/title/tt31710249/
https://social-cdn.vivaldi.net/system/media_attachments/files/114/944/681/947/063/883/original/05ae1bbfa6358b5b.png

2025-07-30 19:53:57 UTC

- reply

💰 Supports multiple dimensions and quantization options - binary 512d version outperforms OpenAI-v3-large while reducing vector database costs by 99.48%
🔍 Processes entire documents in single pass to generate chunk embeddings enriched with document-level context
🎯 Less sensitive to chunking strategies compared to traditional context-agnostic embedding models
https://blog.voyageai.com/2025/07/23/voyage-context-3/

2025-07-29 21:44:02 UTC

#Neko Self-Hosted Virtual #Browser #opensource #docker #webrtc #privacy #collaboration 🐱
🌐 Run fully-functional browser in virtual environment with #WebRTC streaming technology
🔒 Secure and isolated browsing environment accessible from anywhere with complete privacy
👥 Multiple users can access simultaneously making it perfect for team collaboration
🎥 Host watch parties and interactive presentations without in-person gatherings
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/114/938/703/773/233/062/original/dbb8d1e1c5891b84.mp4

2025-07-29 21:44:02 UTC

- reply

🛠️ Supports multiple browsers: #Firefox #Chrome #TorBrowser #Edge #Brave and desktop environments
📱 Not limited to browsers - run any Linux application including #VLC media player and full desktop environments
🎯 Use cases include collaborative debugging, teaching, support, persistent workspace and throwaway browsing
🔧 Easy #Docker deployment with comprehensive documentation and room management via neko-rooms
https://github.com/m1k1o/neko

2025-07-29 21:18:17 UTC

- reply

📊 Includes #SAST, #SCA (dependency scanning), and secrets detection capabilities
🔧 Easy installation via #Homebrew, #pip, or #Docker with seamless #GitHub #GitLab integration
https://github.com/semgrep/semgrep

2025-07-29 21:18:16 UTC

#Semgrep static analysis tool for #code scanning at ludicrous speed 🔍
🔍 Supports 30+ languages including #Python #JavaScript #Java #Go #C #Rust #TypeScript #php and more
🛡️ Finds bugs, enforces #security guardrails and coding standards with semantic pattern matching
⚡ Runs locally by default - code never uploaded, works in #IDE, pre-commit hooks & #CI/CD workflows
🤖 #AI-powered #SemgrepAssistant provides intelligent triaging and step-by-step remediation guidance
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/114/938/602/532/027/307/original/32314ecd2d10f422.png

2025-07-29 15:19:02 UTC

- reply

📊 Views encapsulate complex SELECT statements with joins & calculations, appearing as regular tables to #Laravel applications
🚀 Real performance gains: Moving profit margin calculation from accessor to stored generated column achieved 80% speed improvement (350ms to 70ms)
💡 Best for: frequently queried computed values, complex reporting queries, and scenarios where database-level operations are needed

2025-07-29 15:19:01 UTC

#SQL Generated Columns & Views in #Laravel #php #mysql #database #performance 🗄️
🔧 Generated columns automatically calculate values from other columns using #SQL expressions - available as virtual (computed on query) or stored (saved to disk)
⚡ Unlike #Laravel accessors, generated columns enable filtering, sorting & indexing at database level for superior performance
🔍 Example: Convert full_name accessor to generated column using CONCAT(first_name, ' ', last_name) in migrations
🧵 👇
https://social-cdn.vivaldi.net/system/media_attachments/files/114/937/189/939/706/942/original/6a9c7237b35f2a43.png

2025-07-17 15:38:10 UTC

Who do I have to buy a giant coffee for so that this crap that’s been around for three million years finally gets fixed? 🤔 🍵 😏
In #PHP, you can use both elseif and else if - except when you use the alternative control structure syntax (: ... endif;), where ONLY elseif works.
Why does it have to be so inconsistent and confusing?
Wouldn’t it be great if PHP just picked one way to do things?
@ThePHPF 👋
https://social-cdn.vivaldi.net/system/media_attachments/files/114/869/317/287/873/067/original/980c0afbe2b2364f.png

2025-05-27 11:33:37 UTC

No, this is not a screenshot from 1999 where someone tried something with Dreamweaver.... this is now... 🙈
That's what happens when you invest in ads on X instead of investing in yourself.....
#webdesign #ui
https://social-cdn.vivaldi.net/system/media_attachments/files/114/579/578/368/071/345/original/941c426991ac6b22.png

2025-04-05 14:25:40 UTC

- reply

⚖️ Licensed under GNU Affero General Public License Version 3 (#AGPLv3)
https://github.com/govigilant/vigilant

2025-04-05 14:25:40 UTC

- reply

🚨 Customizable notifications to alert you exactly
when and where you need
📊 Built with #Laravel, #Livewire, #TailwindCSS, and #AlpineJS for powerful performance
🛠️ Available as hosted solution or self-hosted option with complete documentation

2025-04-05 14:25:39 UTC

Vigilant: All-in-One #WebMonitoring Tool
Vigilant allows you to monitor all aspects of your website including uptime, #DNS, #Lighthouse, and broken links
🔍 Comprehensive monitoring of website performance, uptime, and DNS configurations
🧵 👇 #monitoring #DevOps #php #Laravel
https://social-cdn.vivaldi.net/system/media_attachments/files/114/285/814/722/209/741/original/10842edb50d499fa.png

2024-12-08 20:04:57 UTC

#Amazon #nova - my very first question I always like to ask 😉
#ai #llm #php #Laravel
https://social-cdn.vivaldi.net/system/media_attachments/files/113/618/995/288/805/336/original/d064cea75ceafd4d.jpeg

2024-12-07 08:51:29 UTC

#Meta's #Llama3 3.3 70B: Advanced #AI Language Model Redefines Performance Standards 🚀
🔹 #Meta's latest #opensource #LLM features 70B parameters and 128K token context window, matching 405B model performance while requiring fewer resources
https://social-cdn.vivaldi.net/system/media_attachments/files/113/610/685/779/170/910/original/3888e57185bfdfd2.jpeg

2024-12-05 23:48:40 UTC

DeepSeek-R1-Lite-Preview is now live: unleashing supercharged reasoning power! 🔍 o1-preview-level performance on AIME & MATH benchmarks. 💡 Transparent thought process in real-time. 🛠️ Open-source models & API coming soon! 🌐
https://chat.deepseek.com/
#ai #llm
https://social-cdn.vivaldi.net/system/media_attachments/files/113/602/889/023/447/875/original/4673527bee45d1af.jpeg

2024-12-05 21:42:25 UTC

Compare Mode in #Google #AI Studio: Your Companion for Choosing the Right Gemini Model 🔍 🤖
🔄 Latest addition to #GoogleAI Studio lets developers test multiple #Gemini and #Gemma models simultaneously
https://aistudio.google.com/app/prompts/new_comparison
#ai #llm
https://social-cdn.vivaldi.net/system/media_attachments/files/113/602/392/361/582/521/original/c41f71e4b60306f4.png

2024-12-05 21:38:19 UTC

#Google #Gemini 1.5 Flash-8B is now production ready 🚀
🔹 Technical Specs: #GoogleDeepMind's latest #LLM maintains performance of original Flash model while being smaller & faster
https://developers.googleblog.com/en/gemini-15-flash-8b-is-now-generally-available-for-use/
#ai #llm

2024-12-05 16:24:06 UTC

The Next Frontier: #SamAltman on the Future of #AI and Society
Altman discusses his corporate strategy at #OpenAI, the transformative potential of AI, and the ethical dilemmas it presents, in an interview at the 2024 #NewYorkTimes Summit.
📺 https://youtu.be/tn0XpTAD_8Q
Summary ↓

2024-12-05 14:00:39 UTC

#OpenAI Announces Major December #AI Launch Series 🎄
🚀 #OpenAI launches "Shipmas": 12 consecutive days of livestreams featuring new #AI releases and updates
📊 #ChatGPT reaches milestone of 300M weekly active users, processing 1B messages daily

2024-12-04 20:53:22 UTC

#Amazon #nova compared to others...
#ai #llm
Source: https://x.com/ArtificialAnlys
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/537/243/617/423/original/31f4e0c3aff712d8.jpeg
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/537/299/665/841/original/078a61e8e41eefaa.jpeg
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/537/347/779/142/original/eb64eb5d881fd60f.jpeg
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/537/401/577/982/original/80e899bbab2c263a.jpeg

2024-12-04 20:20:49 UTC

Hmmm 🤔 maybe it's time to give #AmazonQ a try.
It's nice to see that a company is aware of the fact that there is not only VSCode out there.
#AI #LLM #coding
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/409/108/037/174/original/24d603657925656c.jpeg
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/409/168/075/790/original/bab6627f0a7f2667.jpeg

2024-12-04 19:22:29 UTC

#Amazon Nova: Comprehensive Guide to Next-Gen Foundation Models 🔮
#AI #AWS #AmazonBedrock #LLM #MultiModal #GenerativeAI #MachineLearning #MLOps #CloudComputing
Examples: https://youtu.be/wGeRLCveGtE
Keynote: https://youtu.be/LY7m5LQliAo?t=6315
https://social-cdn.vivaldi.net/system/media_attachments/files/113/596/179/796/632/634/original/e3156104348f8cb0.jpeg

2024-12-03 16:59:08 UTC

#Microsoft Launches New Framework for Efficient AI Model Processing 🔧
https://github.com/microsoft/BitNet
🔋 Remarkable energy savings of up to 82.2% on x86 CPUs and 70% on ARM
⚡️ #CPU performance gains of 6.17x (x86) and 5.07x (ARM) for larger #LLM models
#ai #llm
https://social-cdn.vivaldi.net/system/media_attachments/files/113/589/953/736/756/479/original/a7b19475d7f45232.png
https://social-cdn.vivaldi.net/system/media_attachments/files/113/589/953/812/609/826/original/1655b0e25f32e904.png

2024-12-03 15:19:16 UTC

Is there anyone out there who has build unit tests for #PDF using #ai ???
I'm thinking about using a #Vision #llm to check specifics inside a pdf - so checking if specific elements are there and also at the correct position...
#coding #testing
https://social-cdn.vivaldi.net/system/media_attachments/files/113/589/560/924/984/678/original/29af32c6c810962c.mp4

2024-12-01 12:14:05 UTC

Advanced Video Layer Separation Technology Unveiled 🎥
📺 https://youtu.be/SD-VCNvTBg4
🔬 #Google #DeepMind and #UniversityOfMaryland researchers develop innovative #AI video decomposition system

2024-11-29 13:33:46 UTC

Next Generation Language Model Release Advances Targeted Training 🎯 #ai #llm
🤖 #NexusFlow introduces #AtheneV2 72B parameter model suite, fine-tuned from #Qwen 2.5
📊 Two specialized variants push performance boundaries:
https://social-cdn.vivaldi.net/system/media_attachments/files/113/566/497/058/293/937/original/831806f8095e8692.jpeg

2024-11-27 17:17:17 UTC

Transform Design into Code: What's New in Screenshot-to-Code 🎯
🤖 Powered by #AI models: #ClaudeSonnet 3.5 and #GPT4 for precise code generation
💻 Supports multiple tech stacks:
• #HTML + #Tailwind
• #React & #Vue integration
• #Bootstrap & #Ionic options
• #SVG capabilities
🎥 New feature: Convert video/screen recordings into functional prototypes
🛠️ Development options:
• Try hosted version at https://screenshottocode.com/
• Run locally with #Docker or manual setup
• Requires #OpenAI or #Anthropic API keys
Source: https://github.com/abi/screenshot-to-code

2024-11-27 00:55:34 UTC

Despite all its strengths, I now firmly believe the worst thing #Laravel has done to most developers is this:
It has made people think that #database design is something you do in a text file (migrations). For many developers, this kind of "designing" has become the new standard. That's crazy.
Migrations are not meant to be used for designing. There are so many great visual database design tools where you see the whole picture: foreign keys, relations—everything. But instead, most people stare at a single #PHP file and call that "designing a database." It's insane.
Migrations should be created AFTER designing the whole database. Use proper tools to design it first and then generate migrations! Use fucking #ai to create this stupid migration files!
Ahhh, this trend drives me 😡!
https://social-cdn.vivaldi.net/system/media_attachments/files/113/552/186/531/063/945/original/bd74f9edb67e126e.mp4

2024-11-26 14:14:01 UTC

dear #jetbrains users out there 👋
please give me a helping hand and leave a 👍 on this feature request here: https://youtrack.jetbrains.com/issue/LLM-13315
➡️Support #Gemini experimental models
thank you ‼️😘
#ai #llm #coding #phpstorm
https://social-cdn.vivaldi.net/system/media_attachments/files/113/549/666/937/326/349/original/bb346d12d7c6ffa6.png

2024-11-26 11:05:49 UTC

Advanced Reasoning Model: #ai #llm Marco-o1 Pushes Boundaries in Problem-Solving 🧠
🔬 Built on #Qwen2, focusing on open-ended reasoning beyond traditional tasks
💡 Key Innovations:
🤔 #ChainOfThought fine-tuning for structured reasoning
🌳 Monte Carlo Tree Search (#MCTS) for solution space exploration
🔄 Novel reflection mechanisms for self-improvement
🎯 Multiple action granularities for complex problem-solving
📊 Performance Highlights:
📈 +6.17% accuracy on MGSM English dataset
📈 +5.60% accuracy on MGSM Chinese dataset
🌐 Excels in translation tasks, especially with colloquial expressions
🛠️ Technical Features:
• Fine-tuned on 60,266 training samples
• Implements step & mini-step MCTS strategies
• Utilizes confidence scoring for path selection
• Incorporates self-reflection mechanisms
⚡️ Project Status: Research work in progress with continuous optimization
https://github.com/AIDC-AI/Marco-o1

2024-11-25 23:13:39 UTC

New Standard for Enterprise AI Data Integration 🌐
🔗 #Anthropic launches Model Context Protocol (#MCP), an #opensource standard enabling seamless connection between #AI systems and enterprise data sources
💻 Local MCP server support added to #Claude Desktop with pre-built connectors for #GitHub, #Slack, #GoogleDrive, #Git, and #Postgres
🤝 Early adopters #Block and #Apollo integrate MCP, while #Zed, #Replit, #Codeium, and #Sourcegraph enhance their development platforms
🛠️ Developers can start testing with existing #Claude for Work accounts, access pre-built MCP servers, and contribute to open-source repositories
🔄 Protocol eliminates need for separate connectors, allowing AI systems to maintain context across different tools and datasets
https://www.anthropic.com/news/model-context-protocol
https://social-cdn.vivaldi.net/system/media_attachments/files/113/546/127/898/113/427/original/a013436be9161038.jpg