AlphaGo vs ChatGPT: The Future of Wealth Management

Chinese AI stocks are absolutely soaring, but have you really done your homework on *where* to place your bets? Let's briefly discuss what's gaining traction right now. Beyond ChatGPT and Gemini, there was one app that seemed to have the potential to challenge the dominance of the Magnificent 7: Deepseek. However, its popularity is rapidly declining. A closer look reveals it's lost over half its user base in just one year. My best guess? Likely privacy concerns and intense competition. This phenomenon begs a critical question for any business navigating the AI investment landscape: **Are we truly evaluating the sustainable competitive advantages and potential pitfalls of these rapidly emerging AI players, or are we just riding the hype cycle?** There's a lot more to uncover about the intriguing world of Chinese AI. For further insights, I encourage you to explore this great resource: [https://lnkd.in/gJc5KkSa

Assistants like ChatGPT, Gemini, and Perplexity don’t show fixed results—they generate answers that vary with every run, every model, and every user. “AI rank tracking” is a misnomer—you can’t track AI like you do traditional search. But that doesn’t mean… Read more ›

I've been asking people at various meetups this question. "When will we move beyond the current AI hype to a deeper understanding of what the real impact of AI will be in the next 10 years?" The hype (both negative and positive) is clouding thinking about how to use this very useful tool. Gathered some thoughts in the post below. #AI #LLM #AGI #ML

How Investors Are Using ChatGPT to Outperform Traditional Methods. AI isn’t just for conversations anymore, it’s rewriting the rules of investing. Traders and investors are now leveraging ChatGPT to analyze markets, spot opportunities, and make smarter decisions faster than traditional methods allow. The question is no longer if AI will change trading, but how fast you’ll adapt. Ready to see how ChatGPT is reshaping investment strategies? Read the full article here👇 https://lnkd.in/g_y6qerj #ChatGPT #OpenAI #AI #Leadership

One of two (likely) rants for the day AI will be the death of social media. You can’t believe videos, photos, or even someone’s thoughts. I continue to see post after post clearly designed by AI. No originality and nothing from the person or human behind it. Why follow someone when it’s an AI chat bot in essence posting? Maybe from here on out I’ll only interact with posts that have typos and no — between every phrase. Here is my ChatGPT clean up of this post too 🤮 That’s not say AI and generative tools aren’t amazing. But at what cost - in art, in connection, in literature.

The world of market analysis just got a major upgrade. For years, I've relied on McKinsey & Company's reports to stay ahead of market trends, but the launch of their new AI #chatbot, is a total game-changer. We've all spent hours sifting through information, trying to verify sources. This new tool changes that. It's not just another ChatGPT, picking the date from open web and we checking the source of information; it's a specialized assistant trained on decades of trusted research. This is huge What do you think? Is this the future of business research? #McKinsey #AI #Innovation #MarketTrends #BusinessStrategy #TrustedData #research #marketresearch

Like many I use AI with a certain level of curious skepticism. Through daily experimentation and use I’ve seen its utility but also its deep flaws. We would be foolish to ignore the significance of such a technological advancement. One that will become a powerful tool that can be employed for the positive but also for the negative. This article is an excellent exploration on the inherent limitation and dangers of LLMs and how far the industry still has to go. The reality of what AI will ultimately become remains to be seen and is hard to predict because of the current bubble we are in. My hedge is that it will fall far short of the overhyped grandiose claims of AGI/ASI. https://lnkd.in/eXgBcT5a

Please stop mixing up goals in a single AI product — keep it focused on a very specific task. Trust me, you will save time, cost, and eventually end up with something you’ll actually use efficiently every day. From my experience in the market, I see people missing a very simple point about AI usage. In simple terms, what an LLM is: it generates text based on the context you give it — that’s it. Think of it like a brain, a brain that can process words or thoughts within a single context, like a dummy person named Mr. “X.” Now imagine giving this one person three very different tasks. Over time, his performance will drop, because he’s stretched thin across multiple goals. Instead, split the work. If possible, assign those three tasks to three separate people. This reduces complexity and keeps each task aligned to one clear purpose. I’m not saying the work must be done in just one step or one layer. It can have multiple layers — but all directed toward a single, well-defined goal. You’re not building ChatGPT with Agent Mode, image generation, and more. The more domains you pile into one system, the harder it becomes for the LLM to complete the task effectively. Please understand: a well-researched architecture and solid evaluation are what actually lead to meaningful, efficient products. I believe in the near future we’ll accept this truth — that investigating the problem statement matters more than spending thousands on products that sound cool but are difficult for end users to use. AI is amazing. But using it efficiently within a particular domain will make your work and projects far superior to those who misuse it. If you mix up domains and expect a single solution for everything, you risk disappointing users, blocking their simple expectations, and missing out on what could have been the best deal.

The truth about AI's knowledge might surprise you. A recent Semrush study (June 2025) reveals something fascinating about where ChatGPT sources its information. The largest contributor? Reddit, making up 40.1% of all citations. Think about that for a moment. The world's most talked-about AI is primarily learning from community discussions and internet debates. Wikipedia follows at 26.3%, with YouTube and Google making up much of the remainder. This raises important questions about AI's knowledge foundation. When we ask ChatGPT for answers, we're essentially getting a refined version of crowd-sourced wisdom - from Reddit threads to wiki entries. It's both intriguing and concerning how this circular flow of information works: Reddit shapes AI, and AI increasingly shapes our understanding. Worth considering as we navigate the AI revolution. Does this change your trust in AI?

We really should stop calling it “AI.” The “A” is right. The “I” is not. I love these tools. I use them every day as a force multiplier. But “intelligent”? Not at all. Here’s the truth: they’re large language models (LLMs). - They try to understand what you say in context. - They use a massive scale of data and parameters. - They generate a response that sounds like what you want to hear. That’s prediction at scale, not intelligence. And here’s the rub: LLMs can only pull from what’s already out there, everything from peer-reviewed studies to Wikipedia entries or Reddit posts. So treating ChatGPT (or Gemini, or Copilot) as a source of truth is like: - Treating Wikipedia as a primary reference. - Treating Reddit as market research. Useful? Absolutely. Enough on its own? Absolutely not. The intelligent one remains you. These tools aren’t replacing us, they’re multiplying us. Like electricity, they’ll eventually run through everything. But they’re not here to think for us. 👉 Do you think calling it “AI” is setting the wrong expectations? 💡 I’ve unpacked more of this in my latest blog post: https://lnkd.in/eKnU4ZEf #ArtificialIntelligence #LLM #FutureOfWork #Leadership #BrightGrain #GrowDifferent