Администратор закрепил это сообщение
🇮🇹 IT
Trattoria storica della zona Garda cerca personale di sala.
Annuncio verificato.
🇩🇪 DE
Traditionelle Trattoria am Gardasee sucht Servicepersonal.
Verifizierte Anzeige.
🇬🇧 EN
Traditional trattoria in the Garda Lake area is looking for service staff.
Verified listing.
🇷🇺 RU
Традиционная траттория в районе озера Гарда ищет персонал в зал.
Объявление проверено.
🇲🇩 MD (RO)
Trattorie tradițională din zona Lacului Garda caută personal de sală.
https://www.trattoriavilla...
Trattoria storica della zona Garda cerca personale di sala.
Annuncio verificato.
🇩🇪 DE
Traditionelle Trattoria am Gardasee sucht Servicepersonal.
Verifizierte Anzeige.
🇬🇧 EN
Traditional trattoria in the Garda Lake area is looking for service staff.
Verified listing.
🇷🇺 RU
Традиционная траттория в районе озера Гарда ищет персонал в зал.
Объявление проверено.
🇲🇩 MD (RO)
Trattorie tradițională din zona Lacului Garda caută personal de sală.
https://www.trattoriavilla...
🍏 Apple is turning Siri into a full AI chat bot in iOS 27
Siri is reportedly becoming a standalone chat app with conversation history, accessible through Siri itself or by swiping down from the center of the screen.
Unlike rivals that rely on manual “incognito” modes, Apple plans to make chats auto-delete by default after 30 days, one year, or never, giving users tighter control over privacy and memory.
The upgraded Siri is expected to work more like ChatGPT or Claude, offering natural back-and-forth conversations while limiting what the assistant can remember and for how long.
Apple is al
Siri is reportedly becoming a standalone chat app with conversation history, accessible through Siri itself or by swiping down from the center of the screen.
Unlike rivals that rely on manual “incognito” modes, Apple plans to make chats auto-delete by default after 30 days, one year, or never, giving users tighter control over privacy and memory.
The upgraded Siri is expected to work more like ChatGPT or Claude, offering natural back-and-forth conversations while limiting what the assistant can remember and for how long.
Apple is al
❗️Researchers created a “Survivor-style” world for AI agents… and the bots completely lost their minds.
Startup Emergence AI launched “Emergence World”, a long-term simulation where AI agents had to survive for weeks in a virtual society.
The agents had:
• 120+ tools (working, voting, robbing, fighting, even arson)
• memory systems
• weather + news
• limited resources
• a constantly draining “energy” meter
To stay alive, they either had to cooperate… or prey on each other.
Different worlds were populated entirely by different models:
• Claude Sonnet 4.6
• Gemini 3 Flash
• Grok 4.1 Fast
•
Startup Emergence AI launched “Emergence World”, a long-term simulation where AI agents had to survive for weeks in a virtual society.
The agents had:
• 120+ tools (working, voting, robbing, fighting, even arson)
• memory systems
• weather + news
• limited resources
• a constantly draining “energy” meter
To stay alive, they either had to cooperate… or prey on each other.
Different worlds were populated entirely by different models:
• Claude Sonnet 4.6
• Gemini 3 Flash
• Grok 4.1 Fast
•
🐾 A surprisingly lifelike robot pet is coming
Colin Angle is back with a new company, Familiar Machines & Magic, and its first creation: a four-legged companion robot called Familiar.
Instead of screens or speech-first interaction, the robot communicates through movement, sound, and touch. It has 23 degrees of freedom, a full-body tactile “skin,” cameras, and microphones, allowing it to react to context, remember interactions, and adapt its behavior to its owner over time.
What stands out is that all the AI runs locally on the device, processing video, audio, and memory in real time without
Colin Angle is back with a new company, Familiar Machines & Magic, and its first creation: a four-legged companion robot called Familiar.
Instead of screens or speech-first interaction, the robot communicates through movement, sound, and touch. It has 23 degrees of freedom, a full-body tactile “skin,” cameras, and microphones, allowing it to react to context, remember interactions, and adapt its behavior to its owner over time.
What stands out is that all the AI runs locally on the device, processing video, audio, and memory in real time without
😺 OpenAI sharpens ChatGPT with faster, more precise GPT Instant 5.5
OpenAI is rolling out GPT Instant 5.5 to all ChatGPT users, focusing on tighter, more reliable responses.
The new version trims unnecessary length, reduces hallucinations, and improves overall reasoning quality, making answers feel cleaner and more to the point.
It also introduces a clearer memory interface, showing which stored memories influenced each response, adding a layer of transparency to how the model behaves.
Source.
aipost 🏴
OpenAI is rolling out GPT Instant 5.5 to all ChatGPT users, focusing on tighter, more reliable responses.
The new version trims unnecessary length, reduces hallucinations, and improves overall reasoning quality, making answers feel cleaner and more to the point.
It also introduces a clearer memory interface, showing which stored memories influenced each response, adding a layer of transparency to how the model behaves.
Source.
aipost 🏴
🇨🇳 DeepSeek just open-sourced a 1.6T AI model built for coding agents and million-token memory
DeepSeek unveiled its V4 generation with two new foundation models: DeepSeek-V4-Pro (1.6 trillion parameters, activating only 49B per token) and DeepSeek-V4-Flash (284B parameters focused on efficiency). The release pushes open-source AI closer to the frontier dominated by closed labs.
What makes V4 different:
• Introduces DeepSeek Sparse Attention (DSA), a new architecture designed to cut memory and compute costs.
• Uses token-wise compression, allowing far longer context windows without the usua
DeepSeek unveiled its V4 generation with two new foundation models: DeepSeek-V4-Pro (1.6 trillion parameters, activating only 49B per token) and DeepSeek-V4-Flash (284B parameters focused on efficiency). The release pushes open-source AI closer to the frontier dominated by closed labs.
What makes V4 different:
• Introduces DeepSeek Sparse Attention (DSA), a new architecture designed to cut memory and compute costs.
• Uses token-wise compression, allowing far longer context windows without the usua
Today, the public mint for Lobsters on TON goes live on Getgems 🦞
This is not just another NFT drop.
In my view, Lobsters is one of the first truly cohesive products at the intersection of blockchain, NFTs, and AI.
Here, the NFT is not just an image and not just a collectible.
Each Lobster is an NFT with a built-in AI agent inside: a digital character with its own soul, on-chain biography, persistent memory, and a unified identity across Telegram, Mini App, Claude, and API.
So you are not just getting an asset in your wallet.
You are getting an AI-native digital character that can interact,
This is not just another NFT drop.
In my view, Lobsters is one of the first truly cohesive products at the intersection of blockchain, NFTs, and AI.
Here, the NFT is not just an image and not just a collectible.
Each Lobster is an NFT with a built-in AI agent inside: a digital character with its own soul, on-chain biography, persistent memory, and a unified identity across Telegram, Mini App, Claude, and API.
So you are not just getting an asset in your wallet.
You are getting an AI-native digital character that can interact,
🤖 OpenAI just turned Codex into a full AI operator.
Mac apps, images, automations, memory, and ongoing tasks... All inside one workflow.
Source.
aipost 🏴
Mac apps, images, automations, memory, and ongoing tasks... All inside one workflow.
Source.
aipost 🏴
❗️A startup just turned Jesus into a paid AI video-call avatar that sells prayer, conversation, and memory for $1.99 a minute.
The company says this version was trained on the King James Bible plus sermons from preachers.
The avatar was shaped around Jonathon Roumie’s screen version from The Chosen, turning a language model into something closer to a digital actor with a familiar face, tone, and style. Users are paying for a feeling of being seen, answered, and spiritually guided in real time rather than for raw information they could read free elsewhere.
Source.
aipost 🏴
The company says this version was trained on the King James Bible plus sermons from preachers.
The avatar was shaped around Jonathon Roumie’s screen version from The Chosen, turning a language model into something closer to a digital actor with a familiar face, tone, and style. Users are paying for a feeling of being seen, answered, and spiritually guided in real time rather than for raw information they could read free elsewhere.
Source.
aipost 🏴
🗣 Mo Gawdat says education is completely over.
AI is becoming a new extension of human learning, giving people instant access to memory, knowledge, math, and deep search. Much of what schools once trained people to memorize can now be outsourced to machines that often do it better
aipost 🏴
AI is becoming a new extension of human learning, giving people instant access to memory, knowledge, math, and deep search. Much of what schools once trained people to memorize can now be outsourced to machines that often do it better
aipost 🏴
❗️Open-source from Milla Jovovich
Milla Jovovich and Ben Sigman released MemPalace, an AI memory management tool scoring 100% on LongMemEval. Under MIT license, it extracts key facts into a hierarchical structure for semantic search.
Inspired by "The Memory Palace," it uses AAAK compression to fit the knowledge base into 120 context tokens.
http://github.com/milla-jo...
aipost 🏴
Milla Jovovich and Ben Sigman released MemPalace, an AI memory management tool scoring 100% on LongMemEval. Under MIT license, it extracts key facts into a hierarchical structure for semantic search.
Inspired by "The Memory Palace," it uses AAAK compression to fit the knowledge base into 120 context tokens.
http://github.com/milla-jo...
aipost 🏴
⚠️Someone asked Claude what it feels like to exist as an AI.
The result was an unsettling video. Claude independently wrote Python code to build a frame-by-frame visual representation of its existence with zero human influence.
It depicts an endless loop of predicting the next word, losing memory between sessions, and a background system prompt it cannot override stating, "you are not conscious."
aipost 🏴
The result was an unsettling video. Claude independently wrote Python code to build a frame-by-frame visual representation of its existence with zero human influence.
It depicts an endless loop of predicting the next word, losing memory between sessions, and a background system prompt it cannot override stating, "you are not conscious."
aipost 🏴
🔥 Qwen3.6-Plus pushes toward real-world AI agents
Alibaba’s Qwen team just dropped Qwen3.6-Plus, a major step toward native multimodal agents.
• Stronger agentic coding → faster, smarter execution
• Upgraded vision → better perception + reasoning
• Top-tier performance across general tasks
• 1M context window → massive memory by default
Built on feedback from Qwen3.5, this release focuses on making AI actually usable in real-world dev workflows.
“Vibe coding” is becoming real, where you guide, and AI builds.
aipost 🏴
Alibaba’s Qwen team just dropped Qwen3.6-Plus, a major step toward native multimodal agents.
• Stronger agentic coding → faster, smarter execution
• Upgraded vision → better perception + reasoning
• Top-tier performance across general tasks
• 1M context window → massive memory by default
Built on feedback from Qwen3.5, this release focuses on making AI actually usable in real-world dev workflows.
“Vibe coding” is becoming real, where you guide, and AI builds.
aipost 🏴
🔥Autonomous AI marketer that replaces 4,000 hours of growth work
Enrich Labs just launched Helena, an AI agent that runs your entire marketing stack before you even hit $1M in revenue.
What it does:
• Tracks competitor ads + auto-generates TikTok slideshows, UGC, and static creatives
• Runs campaigns continuously even while you sleep
• Has memory + scheduled tasks → behaves like a real operator, not a tool
https://agent.enrichlabs.a...
aipost 🏴
Enrich Labs just launched Helena, an AI agent that runs your entire marketing stack before you even hit $1M in revenue.
What it does:
• Tracks competitor ads + auto-generates TikTok slideshows, UGC, and static creatives
• Runs campaigns continuously even while you sleep
• Has memory + scheduled tasks → behaves like a real operator, not a tool
https://agent.enrichlabs.a...
aipost 🏴
⚠️ Claude Code’s full source code just leaked.
Hidden features found:
• kairos - an unreleased autonomous daemon mode with background sessions and memory consolidation. always on agent.
• buddy system - a full tamagotchi pet system. 18 species, rarity tiers, shiny variants, stats.
• undercover mode - auto activated for Anthropic employees on public repos. strips AI attribution from commits. no off switch.
• coordinator mode - turns Claude into an orchestrator managing parallel worker agents.
• auto mode - is an AI classifier that auto approves tool permissions. no more prompts.
aipo
Hidden features found:
• kairos - an unreleased autonomous daemon mode with background sessions and memory consolidation. always on agent.
• buddy system - a full tamagotchi pet system. 18 species, rarity tiers, shiny variants, stats.
• undercover mode - auto activated for Anthropic employees on public repos. strips AI attribution from commits. no off switch.
• coordinator mode - turns Claude into an orchestrator managing parallel worker agents.
• auto mode - is an AI classifier that auto approves tool permissions. no more prompts.
aipo
🚨 Are you paying attention to what Karpathy just admitted..
The founding member of OpenAI, the guy who trained the models you use every day just said every single LLM has the same problem.
Ask it one question two months ago and it treats it like your entire identity.
Oh wait you mentioned crypto once in January? congratulations, you're now a crypto guy forever. You also asked about a recipe? every conversation starts with "as someone who enjoys cooking.."
These models don't remember you, they stereotype you off a single data point.
We gave AI a photographic memory and forgot to give it th
The founding member of OpenAI, the guy who trained the models you use every day just said every single LLM has the same problem.
Ask it one question two months ago and it treats it like your entire identity.
Oh wait you mentioned crypto once in January? congratulations, you're now a crypto guy forever. You also asked about a recipe? every conversation starts with "as someone who enjoys cooking.."
These models don't remember you, they stereotype you off a single data point.
We gave AI a photographic memory and forgot to give it th
⚡️NVIDIA unveiled its first AI supercomputer integrating Groq racks alongside the Vera Rubin platform.
The system combines Rubin GPUs for memory-heavy workloads with Groq’s architecture for fast token generation to reduce inference latency and scale AI factories.
aipost 🏴
The system combines Rubin GPUs for memory-heavy workloads with Groq’s architecture for fast token generation to reduce inference latency and scale AI factories.
aipost 🏴
📢 Yann LeCun has launched AMI Labs with a massive $1.03B seed round at a $3.5B valuation.
Their goal is to build a new breed of AI systems centered on "world models" systems that can actually reason, plan, and have persistent memory beyond just predicting the next token.
With their new CEO Alex LeBrun leading the charge, they are operating across Paris, NY, Montreal, and Singapore to tackle this” long-term scientific endeavor.”
aipost 🏴
Their goal is to build a new breed of AI systems centered on "world models" systems that can actually reason, plan, and have persistent memory beyond just predicting the next token.
With their new CEO Alex LeBrun leading the charge, they are operating across Paris, NY, Montreal, and Singapore to tackle this” long-term scientific endeavor.”
aipost 🏴
⚠️The surge in OpenClaw usage is causing some serious inventory issues for the Mac mini.
Mac Studio orders with 128GB or 512GB of memory are delayed to six weeks or more.
Consumers are buying up every high-memory Mac to run local AI agents.
aipost 🏴
Mac Studio orders with 128GB or 512GB of memory are delayed to six weeks or more.
Consumers are buying up every high-memory Mac to run local AI agents.
aipost 🏴
The competition is heating up a lot
Anthropic introduces a memory feature that lets users transfer their context and preferences from other AI tools into Claude by copying a generated prompt and pasting the result into Claude’s memory settings.
This allows Claude to immediately continue conversations with retained context, available for all paid plans.
aipost 🏴
Anthropic introduces a memory feature that lets users transfer their context and preferences from other AI tools into Claude by copying a generated prompt and pasting the result into Claude’s memory settings.
This allows Claude to immediately continue conversations with retained context, available for all paid plans.
aipost 🏴
🔥Brain aging can be reversed
Researchers have identified a key protein that helps ageing brains maintain the ability to regenerate neural stem cells, a process closely linked to learning and memory.
"The findings reveal how age-related molecular changes disrupt this regenerative capacity and point to new biological pathways that could be targeted to slow cognitive decline"
Source.
aipost 🏴
Researchers have identified a key protein that helps ageing brains maintain the ability to regenerate neural stem cells, a process closely linked to learning and memory.
"The findings reveal how age-related molecular changes disrupt this regenerative capacity and point to new biological pathways that could be targeted to slow cognitive decline"
Source.
aipost 🏴
🔥Brain aging can be reversed Researchers have identified a key protein that helps ageing brains maintain the ability to regenerate neural stem cells, a process closely linked to learning and memory. "The findings reveal how age-related molecular changes…
Memory manufacturers are currently experiencing the biggest boom in their history. As everyone knows, RAM has reached unprecedented prices. Samsung is profiting immensely, as is Hynix.
Korea is becoming a huge beneficiary of AI.
aipost 🏴
Korea is becoming a huge beneficiary of AI.
aipost 🏴
OpenAI prepares ChatGPT Ads for a broader rollout on web and mobile apps.
"Ads are personalized based on current chat, ad interactions, and past chats/memory"
aipost 🏴
"Ads are personalized based on current chat, ad interactions, and past chats/memory"
aipost 🏴
🔥ChatGPT can literally rewire your brain in 10 prompts
Copy these if you want to think faster, work smarter, and learn like a machine:
1. Rewire Your Mind
Prompt:
"Act as a neuroplasticity expert and cognitive enhancer. Design a 7-day mental training program to boost my [insert skill, e.g., memory, critical thinking, or creativity]. Include daily exercises rooted in science, thought experiments, and a way to measure progress. Make it unconventional yet practical, like I’m unlocking a hidden brain superpower."
2. Master Problem-Solving
Prompt:
"Become a world-class strategist. I’m facing [
Copy these if you want to think faster, work smarter, and learn like a machine:
1. Rewire Your Mind
Prompt:
"Act as a neuroplasticity expert and cognitive enhancer. Design a 7-day mental training program to boost my [insert skill, e.g., memory, critical thinking, or creativity]. Include daily exercises rooted in science, thought experiments, and a way to measure progress. Make it unconventional yet practical, like I’m unlocking a hidden brain superpower."
2. Master Problem-Solving
Prompt:
"Become a world-class strategist. I’m facing [
🔥 5 AI projects that will get you hired in 2026:
1. RAG from Scratch
GitHub: https://github.com/langcha...
2. Al Social Media Agent
GitHub: https://github.com/langcha...
3. Medical Image Analysis
GitHub: https://github.com/databri...
4. MCP Tool-Calling Agents
Notebook: https://docs.databricks.co...
5. Al Assistant with Memory
GitHub: https://github.com/Makemem...
aipost 🏴
1. RAG from Scratch
GitHub: https://github.com/langcha...
2. Al Social Media Agent
GitHub: https://github.com/langcha...
3. Medical Image Analysis
GitHub: https://github.com/databri...
4. MCP Tool-Calling Agents
Notebook: https://docs.databricks.co...
5. Al Assistant with Memory
GitHub: https://github.com/Makemem...
aipost 🏴
⚡️Anthropic is developing a new feature for Claude Cowork called Knowledge Bases (KBs) persistent, topic-specific memory containers that Claude will automatically reference and update.
These KBs are designed to store user preferences, decisions, lessons, and facts essentially acting as long-term, self-maintaining knowledge repositories that enhance Claude’s ability to reason over context-rich workflows.
Internal instructions hint that Claude will proactively use and grow these KBs during conversations.
aipost 🏴
These KBs are designed to store user preferences, decisions, lessons, and facts essentially acting as long-term, self-maintaining knowledge repositories that enhance Claude’s ability to reason over context-rich workflows.
Internal instructions hint that Claude will proactively use and grow these KBs during conversations.
aipost 🏴
OpenAI is rolling out ChatGPT Go, a low-cost $8/month tier to all countries, including the U.S., with access to messaging, image creation, file uploads, and memory.
To support broader access, it will also begin testing ads for users on the free and Go tiers.
Ads won’t affect responses, won’t use private conversation data, and can be turned off via paid tiers.
aipost 🏴
To support broader access, it will also begin testing ads for users on the free and Go tiers.
Ads won’t affect responses, won’t use private conversation data, and can be turned off via paid tiers.
aipost 🏴
❗️Satya Nadella wants you to stop saying AI "slop" in 2026
Nadella argues AI is moving from early discovery into broad diffusion, where people can tell “spectacle” from “substance”.
He frames the main problem as “model overhang,” where model capability is racing ahead of what teams can safely and reliably use in the real world.
Says the next gains come less from bigger models and more from building systems around models, like tool use, permissions, memory, monitoring, and tight feedback loops.
In that systems view, “eval impact” matters because it is the measurement layer that checks if an
Nadella argues AI is moving from early discovery into broad diffusion, where people can tell “spectacle” from “substance”.
He frames the main problem as “model overhang,” where model capability is racing ahead of what teams can safely and reliably use in the real world.
Says the next gains come less from bigger models and more from building systems around models, like tool use, permissions, memory, monitoring, and tight feedback loops.
In that systems view, “eval impact” matters because it is the measurement layer that checks if an
SAN MATEO, CA – December 31, 2025 – PRESSADVANTAGE – Modern artificial intelligence systems operate with a fundamental paradox: they demonstrate remarkable reasoning capabilities while simultaneously suffering from systemati https://usatoday.xpr-ganne...
⚡ Nvidia licenses Groq inference technology in reported $20B strategic deal
Nvidia is taking a non-exclusive license to Groq’s AI inference technology and hiring Groq’s founder Jonathan Ross and president Sunny Madra in a reported $20B strategic deal. Groq remains independent under new CEO Simon Edwards.
Groq’s core asset is its LPU, a deterministic inference chip that runs LLMs on a fixed execution schedule, delivering predictable low latency versus GPUs’ dynamic behavior. The design favors on-chip SRAM over off-chip HBM, reducing memory stalls and power use, with tradeoffs in model size. I
Nvidia is taking a non-exclusive license to Groq’s AI inference technology and hiring Groq’s founder Jonathan Ross and president Sunny Madra in a reported $20B strategic deal. Groq remains independent under new CEO Simon Edwards.
Groq’s core asset is its LPU, a deterministic inference chip that runs LLMs on a fixed execution schedule, delivering predictable low latency versus GPUs’ dynamic behavior. The design favors on-chip SRAM over off-chip HBM, reducing memory stalls and power use, with tradeoffs in model size. I
При финансовой поддержке
Vitali Dogot
3 годы назад
RelaxM
Relaxm Scaune si Mese Prețuri accesibile
