SK Group Chairman on Early AI Commitment

Three years before ChatGPT launched, Tae-won (Tony) Chey started exploring AI as a new growth direction for SK Group. Speaking with Daniel Newman at Nasdaq, the SK Group Chairman explained his hedge: "That was actually, well, it was about 2019. So I decided that, well, we have to pursue these AI things. At that time, it was not a memory business. The memory chip and AI were too far apart. But the company is called SK Telecom, and the business just stopped there. But I knew that there was something better, a new business for the telecom. And I actually picked the AI side." SK Group built a large internal organization around that bet, before generative AI entered the mainstream. As Chey put it, the arrival of mainstream LLMs like ChatGPT roughly three years later proved the environment he'd already been studying.

🚀 The AI Blackout: What the ChatGPT Outage Reveals About the Fragility of Our New Digital Infrastructure OpenAI confirms a massive service disruption to ChatGPT, triggering a global ripple effect across professional and automated workflows. As the world's most influential AI goes offline, the industry fa... 🔗 Read more: https://lnkd.in/gumXMP26 #AutoKeren #AI #Automation #TechStartup

State of ChatGPT plus as of Today in understanding Indic currency! This is the state of AI models in calculation. Now all of you verify if this cal is true/false

For basic tasks, I'm seeing Claude and ChatGPT run 50-100% slower, use ~3x more tokens, and produce only marginally better output than earlier models. If you're on usage-based pricing, that's a real cost curve to watch. Curious whether others are seeing the same trends — and how you're thinking about usage as these models get more expensive to run at scale. Question: How will end customers ultimately hold these AI vendors accountable for outcome quality rather than effort (a black box)?

A small business nobody had heard of is now recommended by ChatGPT, Gemini, and Claude. They didn't have a massive budget. One strategic shift changed everything. What made AI systems cite them? Here's the breakdown. #AISearch #GenerativeEngineOptimization #SmallBusiness #AIMarketing

ChatGPT wasn't the first AI model. It was the first to make conversation feel like the most natural way to use a computer. By removing friction instead of adding features, OpenAI transformed a research breakthrough into a product millions could understand within seconds. The interface wasn't the innovation. The simplicity of the interaction was. Products change categories. Interfaces change behaviour. Launch Rating: 9.5 / 10 Full breakdown linked in comments.

Artificial intelligence systems like ChatGPT can lie, distort, and withhold information. This is disastrous because they’re rapidly becoming humanity’s most widely used source of information—as super-convenient “truth machines.” “AlterAI” is taking a stand against this. Is it really the only objective chatbot? Learn more here >> https://lnkd.in/g4y7ri3Y Dr. Harald Wiesendanger Stiftung Auswege

3 AI Agent Frustrations Every Founder Running on Claude, ChatGPT, or Gemini Will Recognize Immediately A quick rundown of what's still broken with agentic AI tools, straight from months of daily use.

Behind the Product | ChatGPT Great products are not built overnight. Behind ChatGPT lies years of research, experimentation, engineering excellence, and the collective effort of a multidisciplinary team working to make AI more useful, safer, and accessible. At VOC Media, our Product Stories go beyond features and specifications. We explore: • The idea behind the product • The problem it set out to solve • The people behind the innovation • The technology powering it • Its business and global impact Our first Product Story explores the journey of ChatGPT—one of the most influential AI products of our time. Read the full story on VOC Media. https://lnkd.in/g7SF3VYQ #ProductStory #ChatGPT #OpenAI #ArtificialIntelligence #Innovation #Business #Technology #VOCMedia

AI doesn't have a memory problem. It has a portability problem — tell ChatGPT everything, and Claude still starts from zero. Every tool remembers you a little. None of them share it. So you rebuild the same context in every tab: your stack, your preferences, how you like things written. The models got smarter. Their memory of you stayed stuck in whatever app you typed it into. So, I built Menda. Try it free: https://lnkd.in/gW46N_m6 It's a memory layer that sits underneath the AI you already use. It remembers who you are — your context, your preferences, the facts that matter — once. Then it hands that context to whatever you connect: ChatGPT, Claude, Cursor, or your own agent. Not another chat app. The layer beneath them. → Recall by meaning, not keywords → Any model provider, connect via API or MCP → Your data stays yours — export or delete anything, nothing trained on it Free tier's live: 50 memories, no card. Tell it once, and every tool you use finally starts from something instead of nothing.