| Preparing for Smarter, Faster, Stronger AI Today's AI systems are cool, capable, and, yes, smart. But they often lack the accuracy and depth of knowledge that a given subject's expert would have. That makes AI useful for some more trivial things, while not terribly useful for truly value-additive tasks. Think ChatGPT-4o, the latest version of OpenAI's popular chatbot. It is fast, has a wide breadth of knowledge and can be used to help with a number of different things. Looking for a place to eat? Need someone to edit an email you just drafted? How about a quick refresher on how stocks do in election years? ChatGPT has you covered. Now, if you're looking for, say, an in-depth dive into how quantum computing works, recent industry developments, timely academic insights, etc. – well, that's something where even the latest version of ChatGPT will fall short. But its next iteration won't… At least, that's what OpenAI's chief technology officer, Mira Murati, thinks. She recently said that ChatGPT-3 – the bot's previous version – possessed toddler-level intelligence, while ChatGPT-4 possesses high-school-level intelligence. And she expects that ChatGPT-5 will have PhD-level intelligence on certain subjects. That's incredible. And what's more, she expects ChatGPT-5 will launch sometime in 2025 or '26.
AI Is Progressing at a Break-Neck Pace The most incredible part about this development? The pace of ChatGPT's progress. Let's assume OpenAI's CTO is right about ChatGPT gaining PhD-level intelligence by 2025 or '26. That would mean that ChatGPT progressed from 'toddler' to 'doctor' in the span of less than five years. It takes humans two decades to achieve that, which means ChatGPT is learning about 4X as fast as we are. And unlike humans, its intelligence is theoretically limitless. Of course, there is only so much knowledge that our brains can retain and only so many processes it can run at any given time. But an AI's database – as well as its processing ability – are capped only by its computing capacity. And if we're able to increase that computational power to near-infinity, the AI has no upper limit on its intelligence. To that end, if ChatGPT reaches PhD-level intelligence by next year… and keeps learning at a rate 4X as fast as humans with no limit to its knowledge… then AI could be smarter than even the smartest human by 2030. And at that point, AI will take over the world.
The Final Word Soon, we'll likely live in a world where AI can write books better than the best authors, market products and services better than the best marketers, create products better than the best product developers – even pick stocks better than the best stock-pickers. Indeed, I'm acutely aware of the existential threat that AI poses to my career. We all should be. In my opinion, it is inevitable. In time, AI will likely do your job better than you. It will certainly do mine better than me. And from that perspective, the only path to true financial freedom is through investing in winning AI stocks. If you can't count on your career to guide you to retirement anymore, then you might as well bet on the stuff that's taking it over. That's my opinion, at least. And it may be a bleak way of looking at things. But the flip side of this coin is that if I'm right, then the companies making next-gen AI will see their stock prices soar more than anyone believes is possible… Which means you could make more money than you ever thought possible, too. Needless to say, with all this change in the air, it is urgent that you plug into the right AI stocks to buy at the current moment. Learn all about our top picks in this space. You can stay up to speed with my latest market analysis by reading my Daily Notes! Check out the latest issue on your Innovation Investor subscriber site. Click here to learn more. Regards, | 
Tidak ada komentar:
Posting Komentar