[ad_1]
While the great crypto collapse may have garnered most of the headlines towards the end of last year, history will likely mark the biggest news in disruptive tech for 2022 to be OpenAI’s launch of ChatGPT in late November. It amassed over 1 million users in just five days right on the heels of its launch of Dall-E in April. In a market that has seen shares plummet, OpenAI has reignited heady valuations.
In this week’s World Reimagined, we will look at the rapidly evolving world of Artificial Intelligence (AI), some of its players, and what companies are poised to gain from a pivotal year where AI hits center stage.
Back in 2018, venture capitalist and computer scientist Kai-Fu Lee told CBS, “I believe [AI] is going to change the world more than anything in the history of mankind. More than electricity.”
For AI to realize its full potential, it needs to be widely available. The launch of ChatGPT, Dall-E and Stable Diffusion in 2022, along with their related applications, make it likely that this is the year Kai-Fu Lee’s prediction becomes real to those outside of AI-focused sectors.
The Different Types of AI Today
AI can be divided into three categories based on capabilities:
- Narrow or weak AI is designed to perform a specific task. This application is becoming increasingly common in our daily lives as machine learning and deep learning techniques improve.
- General or strong AI is designed to perform any intellectual task that a human can. So far, this has proved highly elusive, but we are getting closer as an international AI race becomes the new global arms race.
- Super AI is currently only a hypothetical concept. It is an AI that can perform any task better than a human, is sentient, has emotions and has been used to terrify moviegoers for decades – think Westworld, Blade Runner or Skynet.
Examples of narrow AI include Google’s (GOOG) AlphaGo, IBM’s (IBM) chess-playing supercomputer Deep Blue, Apple’s (AAPL) FaceID unlock, the virtual filters on Snapchat (SNAP) and every writer’s best friend, Grammarly. When I asked ChatGPT if it is a type of narrow AI, this was its response: “Yes, ChatGPT is considered a narrow AI, as it is designed to perform specific tasks and has a limited scope of understanding and capabilities. It is trained on a large dataset of text and can generate human-like responses to text-based prompts, but it does not have the ability to perform a wide range of tasks or understand and process information like a human would.”
AI Goes Mainstream
AI is being used across a wide range of industries, such as retail, agriculture, banking, transportation and even by HR departments. Before its public launch, there was already a legion of third-party services and tools being built on top of OpenAI.
For example, on Dec. 12, 2022, Jay Browder, CEO of DoNotPay, tweeted a screen-capture of his company’s AI talking to Comcast’s (CMCSA) AI-driven chatbot to save one of his company’s engineers $120 a year on their internet bill – an AI-to-AI chat that can save you money! The company plans to take on two speeding ticket cases in court next month, using its AI to tell the defendants what to say.
Other ChatGPT-based applications include QuickVid, which is designed to generate viral YouTube shorts, and MightyGPT, which puts the “superpowers of GPT-3 and ChatGPT” on Meta’s (Meta) WhatsApp and soon on Apple’s (AAPL) iMessage. In October 2022, another OpenAI-based service, Jasper, reached a $1.5 billion valuation just 18 months after it launched, having amassed 100,000 customers, around 75% of whom pay $80 or more monthly for the company’s suite of AI-powered writing templates, striking fear into the hearts of contract writers everywhere.
OpenAI and its ChatGPT aren’t the only games in town. Anthropic, which was started by former OpenAI employees, has received over $700 million in funding and developed an AI system similar to ChatGPT that reportedly has more “grace” than ChatGPT. Its AI, dubbed Claude, is currently accessible only through a Slack integration as part of a closed beta. The tech was created using a technique called “constitutional AI” that aims to provide a principle-based approach to aligning AI systems with human intentions, which basically means it will be more polite and less offensive in its responses than other models. When you think that these AIs learn from the internet, offensive responses aren’t much of a stretch. The company has indicated that it is looking for a valuation of at least $5 billion in its upcoming $200 million round. Sam Bankman-Fried, Caroline Ellison and other ex-FTX colleagues collectively invested over $500 million in Anthropic during prior rounds, so their stakes will likely get sold to the highest bidder soon since their assets have been frozen.
Adept AI Labs is another startup whose founders include former leaders of OpenAI and Google AI. It reportedly recently sought a valuation of around $1 billion following its $250 million valuation during its latest funding round last year. Both Adept and Anthropic are looking for valuations that are hundreds of times their projected topline revenue – heady days are back in parts of the tech sector!
The release of Stable Diffusion in 2022 was not remarkable only because it was open source, but even more amazing was its diminutive size; AI code tends to be bulky. Upon its release, it was already possible to run it on some consumer graphics cards. In less than a month, it had been optimized to the point where it could be run on an iPhone, leading Apple to announce in December that it had released “optimizations to Core ML for Stable Diffusion in MacOS12.1 and iOS 16.2, along with code to get started deploying in Apple Silicon devices.” What is likely the most interesting aspect of this is the privacy angle because any data entered by an Apple user using this tool will stay on the device. It also means that after downloading it, users won’t need an internet connection, opening up opportunities in areas where connectivity is challenging.
Innovation is only going to accelerate. According to Senior Partner & Managing Director at Boston Consulting Group and Global Leader of BCG X Sylvain Duranton, “Despite economic headwinds, 60% of BCG’s recently surveyed companies plan to increase their investments in digital and AI in 2023.” According to IDC Research, spending worldwide by governments and businesses on AI technology will top $500 billion in 2023.
Microsoft Gets Serious
The power of ChatGPT lit up social media, with some hypothesizing that it could mean the end of Google’s search dominance. According to Semafor, Microsoft (MSFT) has reportedly been in talks to invest an additional $10 billion in the owner of ChatGPT in a deal that would value OpenAI at $29 billion. The two have been intimately linked since inception. Three years ago, Microsoft invested $1 billion in a project originally co-founded by Elon Musk and Sam Altman, former Y Combinator president, that became known as OpenAI, in an effort to overtake Amazon (AMZN) and Google in their AI efforts.
ChatGPT’s stunning success means OpenAI needs more of everything, as anyone who has tried to use it can attest. The Microsoft OpenAI deal is a highly unusual structure, according to those claiming to be familiar with the deal. First, OpenAI will pay back its first investors. Then, Microsoft gets 75% of OpenAI’s profits until its principal investment is paid back and 49% of profits after that until it reaches a theoretical cap.
Who’s Winning?
We are still in the early innings of the AI race, but it is already clear that Amazon’s AWS is poised to be a winner as AI requires massive amounts of computing power. For example, the CEO of Stability reported that his company used 256 Nvidia (NVDA) A100s on AWS for 150,000 hours at the cost of $600,000 to train Stable Diffusion. ChatGPT obviously is using Microsoft for computing capacity. Click here to read about chip manufacturers that are poised to benefit in my article last month on Healthcare and AI.
The next phase of AI could also benefit the social media OG Meta, given its massive amounts of data, which is vital food for machine learning, and the reality that its entire business is based on AI that seeks to target users with increasingly personalized content. Its business is basically all about AI experimentation and iteration. Don’t count this behemoth out just yet and keep a close eye on what is happening with MightyGPT on WhatsApp.
Alphabet’s Google claims to have image generation AI that outperforms Dall-E and is rumored to have a conversation chat product that is superior to ChatGPT, but so far, it is all just rumor.
We’ve already discussed Microsoft getting even cozier with OpenAI, but what could be yet another major coup for the tech giant is the rumored incorporation of OpenAI’s AI into its Word, PowerPoint, Excel, Outlook, and other apps. The Information recently reported that the company’s Bing search plans to use OpenAI’s ChatGPT technology. For those really interested, I suggest you ask ChatGPT itself about how Microsoft could use it, if you can find a time when it isn’t already at capacity.
The Bottom Line
The internet shrunk the world, allowing us to transact and communicate, regardless of geography. The smartphone placed the whole of human knowledge in the palm of one’s hand and further expanded our ability to communicate, transact and learn. Buckminster Fuller’s “knowledge doubling curve,” pointed out that until 1900, human knowledge doubled approximately every century. By the end of World War II, knowledge was doubling every 25 years.
Today, on average, human knowledge is estimated to double every 13 months. AI will not just accelerate growth in knowledge but will also give us the ability to use that knowledge in ways that will likely make every other technological advance pale in comparison. From healthcare to energy, construction to the cosmos, we will be able to solve more problems, solve them faster, and solve them more effectively than we could have been dreamed of just a few decades ago.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
[ad_2]
Image and article originally from www.nasdaq.com. Read the original article here.