The battle for AI is heating up, and at stake is a godlike power. OpenAI’s revolutionary GPT4 launch brings us one step closer to AGI. Microsoft declared war on Google with its Bing search ChatGPT launch, pulling away millions of users, forcing Google to respond by launching Bard. Amazon and Meta both revealed their own large language AI models. OpenAI also cut the price of its ChatGPT API by 90%, and locked up partnerships left and right. And a thousand other startups have exploded onto the scene hoping to get a piece of the action. Even Elon is entering the ring.
AI can now pass the US medical licensing exam, and Google Coding Interview for a Level 3 Engineer With a $183K Salary. But it can lie to you very convincingly, even tell you to divorce your wife, or worse still, help invent a new biological weapon of mass destruction.
And although AI will NOT become sentient and kill us terminator style, it does have the potential to takeover millions of jobs it can do better than people, and if it becomes powerful enough, even become an arbiter of truth that has the power to rewrite history. In short, AI is the world’s next great superpower, for better or for worse. It’s Like an information nuclear weapon, and whoever owns the launch codes will take over the world. So who will ultimately win the war for AI? The race for AI supremacy is on, and whoever emerges victorious may well plot our future for decades to come.
AI products like ChatGPT have reached 100 million users in a few months, faster than any product in history. So AI could be bigger than any tech movement that's come before, more on the scale of something like the Industrial Revolution that changed the course of human history. That’s why so many companies are investing billions into this area, even before they’ve made any money. AI is becoming the new semiconductor industry that gave Silicon Valley its name. Companies like Intel invested billions every year in a new fab facility to manufacture its latest generation chip. You had the 1st generation x86 microprocessor, then second generation and so forth. And today, we’ve realized that if you give models 1000 times more data, you can produce something truly remarkable like ChatGPT or MidJourney. So everyone is spending billions to try to ingest yet 1000 more data - and you see companies like OpenAI doing this and then producing new generations GPT1, then GPT2 to GPT4 and so forth over time. Instead of billions being spent on manufacturing facilities for chips, they are being spent on data and data labeling in this arms race to release the next better computing model.
And while Google, Apple, Facebook and others have had an unassailable monopoly for more than a decade, platform shifts like AI present a once in a lifetime opportunity to leapfrog other companies. IBM made all its money on selling computers to companies and was happy to outsource the original micro-computer PC chip and operating system to Intel and Microsoft, until they came to dominate the market. Apple was all but shut out in the PC market, but ended up dominating mobile with the advent of the iPhone. The personal computer IBM and giving rise to massive companies like Amazon and Google. So how will the chips fall this time with AI?
If proprietary data will be key to winning the AI race in the long run, maybe the most interesting dark horse is Quora Poe. ChatGPT has stormed to unprecedented usage by offering a way to ask questions and have direct responses. This is exactly what Quora has been doing since 2009.
Snapchat is even getting in on the action, launching a chatbot powered by ChatGPT called “My AI” as a pin within the app’s chat tab above conversations with friends. It’s initially a fast, mobile-friendly version of ChatGPT inside Snap, except the company’s trained to adhere to Snap’s specific trust and safety guidelines and not give responses that include swearing, violence, or other explicit content. It’s also been modified to decline answering homework related questions, which is functionality that has already gotten ChatGPT banned in some schools
It’s possible when all’s said and done, the biggest winners may actually be the chipmakers Nvidia and TSMC who power all these tech giants building models on top of their infrastructure. Watch the video to find out more!
#aiarmsrace #artificialintelligence #googlebard #chatgpt #openai
