Stay ahead with the latest tech breakthroughs, industry trends, and cutting-edge news on our Innovation Feed channel.
https://samirb.com.np/
https://samirb.com.np/
#MistralAI #Devstral2 #AICoding
Mistral AI surfs vibe coding tailwinds with new coding models
French AI whiz Mistral is dropping Devstral 2 to amp up coding smarts, riding the wave of natural language code tools.
Key highlights
⦁ Devstral 2 packs 123B parameters for heavy-duty coding, while the smaller 24B Devstral Small runs on everyday hardware.
⦁ Mistral Vibe CLI lets devs automate code via chat-like commands, scanning files and Git for smart context.
⦁ Open-source vibes: Devstral 2 under modified MIT (free API now, then $0.40/$2.00 per million tokens); Small on Apache 2.0 at $0.10/$0.30.
⦁ Builds on Mistral 3 family, partnering with Kilo Code and Cline for IDE integration like Zed.
⦁ Valued at €11.7B post-funding, Mistral's chasing giants like Anthropic with enterprise-ready AI.
Read More: Mistral Coding Models
Mistral AI surfs vibe coding tailwinds with new coding models
French AI whiz Mistral is dropping Devstral 2 to amp up coding smarts, riding the wave of natural language code tools.
Key highlights
⦁ Devstral 2 packs 123B parameters for heavy-duty coding, while the smaller 24B Devstral Small runs on everyday hardware.
⦁ Mistral Vibe CLI lets devs automate code via chat-like commands, scanning files and Git for smart context.
⦁ Open-source vibes: Devstral 2 under modified MIT (free API now, then $0.40/$2.00 per million tokens); Small on Apache 2.0 at $0.10/$0.30.
⦁ Builds on Mistral 3 family, partnering with Kilo Code and Cline for IDE integration like Zed.
⦁ Valued at €11.7B post-funding, Mistral's chasing giants like Anthropic with enterprise-ready AI.
Read More: Mistral Coding Models
TechCrunch
Mistral AI surfs vibe-coding tailwinds with new coding models | TechCrunch
French AI startup Mistral today launched Devstral 2, a new generation of its AI model designed for coding, as the company seeks to catch up to bigger AI labs like Anthropic and other coding-focused LLMs.
#LegacyUpdate #MicrosoftArchive #OldWindows
Microsoft Download Center Archive
Legacy Update's archive rescues thousands of deleted Microsoft downloads, letting you access old Windows files safely through the Wayback Machine.
Key highlights
⦁ Preserves pre-2020 downloads for Windows 95 to 7,.NET Framework, Visual C++ Redistributables, DirectX, and tools like Virtual PC.
⦁ Searchable catalog from 2012-2024, including Media Feature Packs for N/KN editions and freeware like SQL Server.
⦁ Warns of security risks in legacy files; recommends using Legacy Update for post-install patches.
⦁ Credits Archive Team and Internet Archive for backups, with no Microsoft affiliation—use at your own risk.
⦁ Supports old PC revival without owning content, via public archives.
Read More: Legacy Update Archive
Microsoft Download Center Archive
Legacy Update's archive rescues thousands of deleted Microsoft downloads, letting you access old Windows files safely through the Wayback Machine.
Key highlights
⦁ Preserves pre-2020 downloads for Windows 95 to 7,.NET Framework, Visual C++ Redistributables, DirectX, and tools like Virtual PC.
⦁ Searchable catalog from 2012-2024, including Media Feature Packs for N/KN editions and freeware like SQL Server.
⦁ Warns of security risks in legacy files; recommends using Legacy Update for post-install patches.
⦁ Credits Archive Team and Internet Archive for backups, with no Microsoft affiliation—use at your own risk.
⦁ Supports old PC revival without owning content, via public archives.
Read More: Legacy Update Archive
Legacy Update
Microsoft Download Center Archive - Legacy Update
#AdobePremiere #YouTubeShorts #MobileEditing
Adobe Premiere's YouTube Shorts editor rolls out, but only on iPhone
Adobe has launched a dedicated YouTube Shorts creation space within its Premiere mobile app, currently available only on iOS.
Key highlights
⦁ The new editor offers exclusive templates, transitions, and effects designed specifically for YouTube Shorts creators.
⦁ Users can create, customize, and reuse original templates tailored for short-form vertical videos.
⦁ The app supports quick and easy export options to publish videos directly to YouTube Shorts.
⦁ Though announced for Android, the YouTube Shorts editor is still in testing and not yet available on that platform.
⦁ This integration streamlines mobile editing, aiming to empower creators with professional tools on the go.
Read More: Adobe Premiere YouTube Shorts
Adobe Premiere's YouTube Shorts editor rolls out, but only on iPhone
Adobe has launched a dedicated YouTube Shorts creation space within its Premiere mobile app, currently available only on iOS.
Key highlights
⦁ The new editor offers exclusive templates, transitions, and effects designed specifically for YouTube Shorts creators.
⦁ Users can create, customize, and reuse original templates tailored for short-form vertical videos.
⦁ The app supports quick and easy export options to publish videos directly to YouTube Shorts.
⦁ Though announced for Android, the YouTube Shorts editor is still in testing and not yet available on that platform.
⦁ This integration streamlines mobile editing, aiming to empower creators with professional tools on the go.
Read More: Adobe Premiere YouTube Shorts
9to5Google
Adobe Premiere's YouTube Shorts editor rolls out, but only on iPhone
Adobe’s new Premiere app for mobile is introducing an editor specifically designed for YouTube Shorts, but it is currently only...
#ChromeAI #GeminiSecurity #AgenticBrowsing
Google Chrome adds new security layer for Gemini AI agentic browsing
Google's rolling out a smart new defense in Chrome to keep Gemini's autonomous web browsing safe from sneaky attacks. 🚀
Key highlights
⦁ Introduces 'User Alignment Critic,' an isolated Gemini model that vets every AI action for safety, blocking risky moves like prompt injections.
⦁ Uses 'Origin Sets' to limit agent access to approved sites only, preventing data leaks across iframes or unrelated pages.
⦁ Adds user oversight for sensitive tasks, like banking logins, pausing for manual confirmation to avoid fraud.
⦁ Includes real-time prompt injection detection and automated red-teaming, plus up to $20K bounties for hackers who spot flaws.
⦁ Builds on Chrome's Safe Browsing with layered protections, making it tougher for scams than rival AI tools.
Read More: Chrome Gemini Security
Google Chrome adds new security layer for Gemini AI agentic browsing
Google's rolling out a smart new defense in Chrome to keep Gemini's autonomous web browsing safe from sneaky attacks. 🚀
Key highlights
⦁ Introduces 'User Alignment Critic,' an isolated Gemini model that vets every AI action for safety, blocking risky moves like prompt injections.
⦁ Uses 'Origin Sets' to limit agent access to approved sites only, preventing data leaks across iframes or unrelated pages.
⦁ Adds user oversight for sensitive tasks, like banking logins, pausing for manual confirmation to avoid fraud.
⦁ Includes real-time prompt injection detection and automated red-teaming, plus up to $20K bounties for hackers who spot flaws.
⦁ Builds on Chrome's Safe Browsing with layered protections, making it tougher for scams than rival AI tools.
Read More: Chrome Gemini Security
#Grok420 #xAI #ElonMusk
Elon Musk Confirms Grok 4.20 Release Timeline: Know Potential Launch Date
Elon Musk just teased the upcoming Grok 4.20 AI model, promising a big leap in capabilities from xAI.
Key highlights
⦁ Musk announced Grok 4.20 will drop in 3-4 weeks, likely early January 2026, building on Grok 4.1's recent rollout.
⦁ The new version aims to rival top models like GPT-5 with enhanced reasoning and reduced biases spotted in predecessors.
⦁ xAI is focusing on making Grok more truthful and fun, integrating it deeper into X's feed ranking.
⦁ Early hints suggest multimodal features, but full details are under wraps until launch.
Read More: Grok 4.20 Details
Elon Musk Confirms Grok 4.20 Release Timeline: Know Potential Launch Date
Elon Musk just teased the upcoming Grok 4.20 AI model, promising a big leap in capabilities from xAI.
Key highlights
⦁ Musk announced Grok 4.20 will drop in 3-4 weeks, likely early January 2026, building on Grok 4.1's recent rollout.
⦁ The new version aims to rival top models like GPT-5 with enhanced reasoning and reduced biases spotted in predecessors.
⦁ xAI is focusing on making Grok more truthful and fun, integrating it deeper into X's feed ranking.
⦁ Early hints suggest multimodal features, but full details are under wraps until launch.
Read More: Grok 4.20 Details
Gadgets 360
Elon Musk Says Grok 4.20 AI Model Could Be Released in a Month
Grok 4.20 will be the second major update for the fourth generation of the Grok AI model.
#MediaDrama #TakeoverBid
Paramount Skydance Launches Hostile Takeover Bid for Warner Bros. Discovery Amid Netflix Deal
Paramount Skydance, led by David Ellison, just dropped a bombshell: a $108.4 billion all-cash hostile bid for Warner Bros. Discovery (WBD) at $30 per share, going straight to shareholders to bypass the board. This comes days after Netflix won the bidding war last week with a $72 billion deal for WBD's studio (Warner Bros.) and streaming (HBO Max) assets at $27.75 per share, leaving out cable networks like CNN and TNT.
Why the drama?
⦁ Paramount argues their offer is simpler, faster (shorter regulatory hurdles, thanks to Trump ties), and preserves WBD as a unified powerhouse vs. Netflix's "volatile" stock-and-cash mix that could face multi-jurisdiction scrutiny.
⦁ It's a push to create a mega-media rival to streaming giants, but success hinges on shareholder buy-in—WBD's board favored Netflix for higher potential value post-cable spin-off.
⦁ Comcast was also in the mix earlier, but Paramount's now framing this as the superior play for long-term stability.
This could shake up Hollywood big time! 😲 What's your take on who should win the WBD prize?
Paramount Skydance Launches Hostile Takeover Bid for Warner Bros. Discovery Amid Netflix Deal
Paramount Skydance, led by David Ellison, just dropped a bombshell: a $108.4 billion all-cash hostile bid for Warner Bros. Discovery (WBD) at $30 per share, going straight to shareholders to bypass the board. This comes days after Netflix won the bidding war last week with a $72 billion deal for WBD's studio (Warner Bros.) and streaming (HBO Max) assets at $27.75 per share, leaving out cable networks like CNN and TNT.
Why the drama?
⦁ Paramount argues their offer is simpler, faster (shorter regulatory hurdles, thanks to Trump ties), and preserves WBD as a unified powerhouse vs. Netflix's "volatile" stock-and-cash mix that could face multi-jurisdiction scrutiny.
⦁ It's a push to create a mega-media rival to streaming giants, but success hinges on shareholder buy-in—WBD's board favored Netflix for higher potential value post-cable spin-off.
⦁ Comcast was also in the mix earlier, but Paramount's now framing this as the superior play for long-term stability.
This could shake up Hollywood big time! 😲 What's your take on who should win the WBD prize?
The Verge
Paramount launches a hostile $108 billion bid to snatch Warner from Netflix
It’s not over yet.
#UKAI #ChatGPT #Locai
A new UK AI says it can beat ChatGPT – we tried it and here's what we found
Locai Labs has launched a British AI that claims to outperform GPT-5, Gemini, and DeepSeek in conversational ability and human preference based on benchmarks.
Key highlights
⦁ Uses "Forget-Me-Nt" tech to self-learn continuously without human trainers, avoiding knowledge loss
⦁ Built as a community AI, leveraging a decentralized blockchain-powered network for scaling without huge data centers
⦁ Currently in early access with text chat, file upload, and web search but lacks features like voice or image creation
⦁ Responses felt accurate and comparable to ChatGPT 3.1/4, but with a more formal tone
⦁ Its success depends on scaling effectively against big players like OpenAI and Google
Read More: Locai, the new British AI challenger
A new UK AI says it can beat ChatGPT – we tried it and here's what we found
Locai Labs has launched a British AI that claims to outperform GPT-5, Gemini, and DeepSeek in conversational ability and human preference based on benchmarks.
Key highlights
⦁ Uses "Forget-Me-Nt" tech to self-learn continuously without human trainers, avoiding knowledge loss
⦁ Built as a community AI, leveraging a decentralized blockchain-powered network for scaling without huge data centers
⦁ Currently in early access with text chat, file upload, and web search but lacks features like voice or image creation
⦁ Responses felt accurate and comparable to ChatGPT 3.1/4, but with a more formal tone
⦁ Its success depends on scaling effectively against big players like OpenAI and Google
Read More: Locai, the new British AI challenger
TechRadar
A new UK AI says it can beat ChatGPT – we tried it and here’s what we found
Sorry, ChatGPT: there's a new polite AI in town (and it takes tea breaks)
#OpenAIGPT5 #CodeRed #AICompetition
OpenAI’s GPT-5.2 ‘code red’ response to Google is coming next week
OpenAI is accelerating its GPT-5.2 release to December 9th, 2025, as a direct counter to Google's Gemini 3 launch—fueled by CEO Sam Altman's "code red" alert on intensifying AI rivalry from Google and Anthropic.
Key highlights:
⦁ Accelerated timeline: Originally mid-December, GPT-5.2 aims to close the performance gap with Gemini 3, which impressed Altman and Elon Musk on leaderboards; focuses on speed, reliability, and customization.
⦁ Internal edge: OpenAI's assessments show the upcoming reasoning model surpassing Gemini 3 in key areas like coding and logic, per sources.
⦁ Broader context: Follows GPT-5.1's November rollout; ties into "Garlic" project, a next-gen LLM for specialized tasks like biomedicine, potentially debuting as GPT-5.2 or 5.5.
⦁ Strategic shift: Emphasizes iterative improvements over flashy features amid competitive pressures; release could slip slightly due to OpenAI's flexible schedules.
Read More: OpenAI’s GPT-5.2 ‘code red’ response to Google is coming next week
Think GPT-5.2 will reclaim the AI throne? What's your bet? 🚀
OpenAI’s GPT-5.2 ‘code red’ response to Google is coming next week
OpenAI is accelerating its GPT-5.2 release to December 9th, 2025, as a direct counter to Google's Gemini 3 launch—fueled by CEO Sam Altman's "code red" alert on intensifying AI rivalry from Google and Anthropic.
Key highlights:
⦁ Accelerated timeline: Originally mid-December, GPT-5.2 aims to close the performance gap with Gemini 3, which impressed Altman and Elon Musk on leaderboards; focuses on speed, reliability, and customization.
⦁ Internal edge: OpenAI's assessments show the upcoming reasoning model surpassing Gemini 3 in key areas like coding and logic, per sources.
⦁ Broader context: Follows GPT-5.1's November rollout; ties into "Garlic" project, a next-gen LLM for specialized tasks like biomedicine, potentially debuting as GPT-5.2 or 5.5.
⦁ Strategic shift: Emphasizes iterative improvements over flashy features amid competitive pressures; release could slip slightly due to OpenAI's flexible schedules.
Read More: OpenAI’s GPT-5.2 ‘code red’ response to Google is coming next week
Think GPT-5.2 will reclaim the AI throne? What's your bet? 🚀
The Verge
OpenAI’s GPT-5.2 ‘code red’ response to Google is coming next week
GPT-5.2 could arrive on December 9th
#ChromeExtensions #AIProductivity #GooglePicks
Google’s 10 best Chrome extensions of 2025 could supercharge your browsing – here’s the full list
Google's end-of-year roundup spotlights AI-powered Chrome extensions that transformed browsing into a smart workspace—perfect for devs and power users tweaking workflows.
Key highlights:
⦁ AI companions: Monica and Sider offer sidebar chats with models like Gemini/Claude for quick searches, translations, and image gen; HARPA AI automates tasks like price tracking.
⦁ Writing & notes: Quillbot polishes text with paraphrasing/grammar; Fireflies.ai and Bluedot transcribe/summarize Meet/Zoom/Teams calls automatically.
⦁ Learning tools: QuestionAI tutors on complex topics; eJOY turns web browsing into language lessons with vocab highlights.
⦁ Extras: Phia finds fashion deals via AI pricing trends; Adobe Photoshop for Chrome (with free trial) handles quick image edits like background removal.
Read More: Google’s 10 best Chrome extensions of 2025
Which one's a must-try for your setup? 🚀
Google’s 10 best Chrome extensions of 2025 could supercharge your browsing – here’s the full list
Google's end-of-year roundup spotlights AI-powered Chrome extensions that transformed browsing into a smart workspace—perfect for devs and power users tweaking workflows.
Key highlights:
⦁ AI companions: Monica and Sider offer sidebar chats with models like Gemini/Claude for quick searches, translations, and image gen; HARPA AI automates tasks like price tracking.
⦁ Writing & notes: Quillbot polishes text with paraphrasing/grammar; Fireflies.ai and Bluedot transcribe/summarize Meet/Zoom/Teams calls automatically.
⦁ Learning tools: QuestionAI tutors on complex topics; eJOY turns web browsing into language lessons with vocab highlights.
⦁ Extras: Phia finds fashion deals via AI pricing trends; Adobe Photoshop for Chrome (with free trial) handles quick image edits like background removal.
Read More: Google’s 10 best Chrome extensions of 2025
Which one's a must-try for your setup? 🚀
TechRadar
These are the 10 best Chrome extensions of 2025, according to Google – and there’s one I definitely recommend
Have you heard of AI?
The channel owner can now choose which admins are allowed to ban users.
Telegram Beta for Android v12.2.8 (6334)
Finally After 69 years ...........
Log in to Telegram via Passkeys Without Codes or Phone Numbers
- Telegram is testing login via Passkeys (no phone number or SMS code).
- Sign-in flow: tap “log in using passkey” → opens password manager → biometric/auth confirms → Telegram uses the signature as login + code.
- Key details:
- Must be pre-configured in Telegram security settings.
- Passkey is stored on the creation device but can sync to other devices via password managers (Google Passwords, iCloud, etc.).
- Benefits:
- Less SMS reliance
- Stronger phishing resistance (stored in secure enclave)
- Faster, biometric login
- Status: under development and not yet fully functional.
Only government-authorized agencies will be allowed to tap phone calls: Home Affairs Adviser.
The Home Affairs Adviser Jahangir Alam Chowdhury stated that only government-authorized agencies will be allowed to intercept phone calls in Bangladesh. He assured that no unauthorized entity will have access to such surveillance. The comment came after a law-and-order meeting at the Secretariat on Tuesday.
https://www.itvbd.com/249087
The Home Affairs Adviser Jahangir Alam Chowdhury stated that only government-authorized agencies will be allowed to intercept phone calls in Bangladesh. He assured that no unauthorized entity will have access to such surveillance. The comment came after a law-and-order meeting at the Secretariat on Tuesday.
https://www.itvbd.com/249087
