Apie šį kanalą
„Turing College“ kanalas nagrinėja, kaip duomenų analizės ir dirbtinio intelekto įgūdžiai įgyjami, pritaikomi darbe ir tampa karjeros dalimi. Viena turinio kryptis skirta besimokantiems bei profesiją…
„Turing College“ kanalas nagrinėja, kaip duomenų analizės ir dirbtinio intelekto įgūdžiai įgyjami, pritaikomi darbe ir tampa karjeros dalimi. Viena turinio kryptis skirta besimokantiems bei profesiją keičiantiems žmonėms, kita – vadovams, svarstantiems, kaip ugdyti savo komandą. Įrankių apžvalgos čia pateikiamos greta organizacijų patirčių ir klausimų apie mokymosi naudą. Dėmesys sutelktas į praktinį santykį su technologijomis: ką verta suprasti pačiam, ko mokyti darbuotojus ir kaip įgūdžius parodyti ieškant darbo.
Turinys aktualus svarstantiems karjeros pokytį, norintiems sustiprinti duomenų analizės pagrindą ar suprasti, kaip DI keičia kasdienes užduotis. Vadovams įdomūs gali būti komandos mokymosi ir organizacinių pokyčių aptarimai, o darbo ieškantiems – įrašai apie kandidatavimą į su DI susijusias pozicijas. Žiūrovas gali palyginti skirtingus mokymosi motyvus ir pamatyti, kaip įgūdžių ugdymas siejamas su konkrečia veikla. Kanale kalbama ir apie Vokietijos mokymosi finansavimo priemonę, todėl dalis turinio skirta labai konkretiems praktiniams klausimams. Bendras atspirties taškas išlieka mokymasis, kurį galima susieti su darbu ir tolesniu profesiniu keliu.
Pokalbiuose su įmonių atstovais keliami klausimai, ar vadovui būtina suprasti DI įrankius ir kodėl organizacija renkasi mokyti esamus darbuotojus. Kitame įraše pristatomas kelias į duomenų analitiką neturint programavimo patirties. Šalia tokių istorijų atsiranda praktinių bandymų, pavyzdžiui, savos užrašų ir darbo organizavimo sistemos kūrimas pagal „Notion“ pavyzdį. Formatai įvairuoja nuo pokalbių bei patirties pristatymų iki įrankių palyginimų ir nuoseklių paaiškinimų. Jie leidžia tą pačią technologijų temą matyti iš besimokančio žmogaus, komandos vadovo ar sprendimą išbandančio kūrėjo perspektyvos.
Naujausias epizodas
Reality of running local models in 2026
Online, you'll find two very different claims about Qwen 3.8 27B, a free, open model you can run on your own laptop. One app says it runs at 144 tokens a second on an M5 Max MacBook, faster than…
Online, you'll find two very different claims about Qwen 3.8 27B, a free, open model you can run on your own laptop. One app says it runs at 144 tokens a second on an M5 Max MacBook, faster than Claude Opus 5.5 in Lukas's tests. A Reddit user with the same chip as Lukas says a simple task takes half an hour or more. Lukas, co-founder and CEO of Turing College, uses Opus every day, so he spent a few days trying to replace it with Qwen, running offline on his MacBook Pro.
In this video, Lukas gives both models the same two jobs and times them. First, each one builds a notes app from a single prompt. Then each one redesigns it to work like Evernote, on desktop and phone, scored by 56 automated checks. He shows why Qwen started out 15 times slower than Opus, the setup fix that cut its first job from nine minutes to under three, and how it finally passed all 56 checks with two cloud AIs planning and checking its work.
Then he covers what could change the answer: a faster Mac, a setup where cloud models plan and Qwen writes the code, and the seven-month gap between open models and the best paid ones.
If you're weighing a local model against your AI subscription, this shows what works on a laptop today, what hardware it needs and where it still falls short.
📚 Helpful resources
FREE AI Benchmark Quiz: check your AI skills in 3 minutes: https://www.turingcollege.com/building-with-ai/quiz?quiz_placement=quiz-section&utm_source=youtube&utm_medium=video&utm_campaign=qwen_local_model
Turing College Blog: deep dives on AI tools, agents and career development: https://www.turingcollege.com/blog?utm_source=youtube&utm_medium=video&utm_campaign=qwen_local_model
👉 Subscribe for more head-to-head tests of AI models and coding agents
⏳ Timestamps
00:00 - Can a free model on a laptop replace Claude Opus 5.5?
01:01 - What a local model is and how far it trails the best models
02:03 - Test 1: plugging Qwen into Claude Code
02:40 - Why Qwen was 15 times slower than Opus
03:10 - The setup fix that cut the job from 9 minutes to under 3
04:00 - LM Studio vs MTPLX: why the engine matters
05:35 - Test 2: an Evernote-style redesign and 56 checks
06:54 - How Qwen got to 56 with help from two cloud AIs
08:12 - Would a faster Mac change the answer?
09:30 - What hardware you need to run Qwen locally
10:29 - When free models could match Opus 5.5
11:16 - Verdict: is Lukas keeping his subscription?
#Qwen #LocalLLM #ClaudeCode #ClaudeOpus #OpenSourceAI #AIEngineering #TuringCollege
YouTube Shorts
AI Insights: Careers and Trends
Online Learning: Productivity and Study Tips
Uploads
Web Development Basics & Career Path
Daugiau šio kanalo epizodų
Inside NASDAQ’s Multimillion-Dollar AI Overhaul | Engineering VP
8.6.2026
What does it really take to roll out an AI overhaul in a high-stakes, heavily regulated industry? When a single engineering error could end up on the global evening news, "vibe coding" just doesn't...
No Coding Experience to Data Analyst in One Year - Here’s How She Did It
7.5.2026
Breaking into a technical field from a completely different background is one of the most common career questions we hear. In this episode, we sit down with Oumi — a former B2B sales professional...
I Rebuilt Notion in 7 Days and Saved $5,040 a Year
30.4.2026
What if the SaaS tools you're paying for, like Notion or Confluence, could be replaced by something you built yourself in a week? That's exactly what I set out to test — using AI...
Moody’s Upskilled 14,000 Employees on AI — Here’s What Happened
22.4.2026
Aukse, Country Manager at Moody's Lithuania, sits down to talk about how one of the world's most trusted risk assessment companies is navigating AI transformation — without compromising the trust they've built over more...
Nano Banana 2 Just Changed AI Image Generation (Review & Comparison)
17.4.2026
Nano Banana 2 dropped, and Google is making some bold promises - better world knowledge, text that finally renders properly, character consistency across generations. I tested all of it and I’ll show you exactly...
Stop Buying AI Tools and Do This Instead - Thermo Fisher’s Director
9.4.2026
Most companies hear “AI transformation” and immediately start shopping for tools. Thermo Fisher heard the same thing and did something harder - they invested in their people first. Evelina has spent over a decade...
Most AI Job Applications Get Ignored. Do This Instead
2.4.2026
Most people send 50, 100, sometimes more applications before getting a single interview. The problem isn’t effort - it’s that AI-generated applications all look identical to the AI doing the screening. That’s the loop...
I Built and Deployed an App in 10 Minutes (Vibe Code Anything!)
24.3.2026
Most people have never heard of Google AI Studio, and given how capable it is - it’s the most underrated vibe coding tool available right now. It’s browser-based, requires zero setup, and lets you...
GPT-5.3 Codex Is The Real Deal! (Every New Feature Tested)
6.3.2026
OpenAI dropped GPT-5.3 Codex on the same day as Claude Opus 4.6 — and for the first time in this race, it feels like a real answer. Computer use performance jumped dramatically, while features...
Why Being ‘Good at Your Job’ Isn’t Enough Anymore – This Is What Matters Now (Hostinger)
5.3.2026
Hostinger shipped two AI products in under eight weeks — no restructuring, no big planning cycles, no slowdown. That kind of speed doesn’t come from process. It comes from people. In this episode, Eglė...
Claude Code's Agent Teams Are Insane (Build Your AI Workforce)
26.2.2026
With the Claude Opus 4.6 release, Anthropic shipped Agent Teams as a new Claude Code feature, and it’s the most production-ready multi-agent implementation available right now. You describe a complex task once, and Claude...
Full OpenClaw Tutorial for Beginners - How to Install & Set Up (Clawdbot)
10.2.2026
OpenClaw is probably the first AI that feels less like a tool and more like a person working for you. It has memory that learns about you over time. It can control a computer...
I Built 2 Apps in 3 Hours With Claude Code 2.0 (Vibe Code Anything!)
4.2.2026
Every AI coding tool promises “just describe what you want and it builds it.” But when you try them, you get code that works technically but fails completely in practice. Claude Code 2.0 broke...
Essential AI Skills For 2026
29.1.2026
The AI skills that actually matter in 2026 have nothing to do with writing prompts. Agent design, context engineering, multimodal thinking - these are what separate people getting hired from people getting left behind....
"AI Literacy Changed The Game" – Nord Security’s Head of Talent Explains
22.1.2026
AI is completely reshaping how we hire, who we hire, and what skills actually matter in 2026. Nord Security went from receiving 60,000 applications per year to over 100,000 after AI entered the market....
How I’d Learn AI in 2026 (If I Had To Start Over)
8.1.2026
Starting AI in 2026 gives you an advantage people who started years ago didn’t have. The tools are better, the resources are clearer, and if you follow the right path - you can skip...
How to Learn AI FAST in 23 Minutes (Full Roadmap for 2026)
1.1.2026
Everyone’s talking about AI, but most people waste months learning tools they never use, watching generic tutorials, and still feeling stuck. You’re not behind yet, and the next 6 months could be your biggest...
I Tried ChatGPT-5.2... (Better Than Gemini 3?)
22.12.2025
OpenAI just launched GPT-5.2 after declaring an internal “code red” over Gemini 3’s success. They’re claiming massive improvements in office productivity, coding, and reasoning. But after spending days testing it against Gemini 3 Pro...
5 ChatGPT Hacks That Will Help You Become a PRO!
18.12.2025
If you’ve ever asked ChatGPT something and got back a vague, copy‑paste answer, you’re not the problem - it’s how you’re using the tool. In this video, you’ll learn a set of simple techniques...
The Truth About Starting an AI Career in 2026
11.12.2025
Headlines say the “AI bubble” is about to burst. Over 150,000 tech jobs were cut in October alone. But the data tells a completely different story... Industries embracing AI are seeing 3x higher revenue...
Innohub TV
Šį turinį žiūrėkite ir Innohub TV
Atrinkome jums Innohub TV klipų, kurie tęsia tą pačią temą.
