And it’s not only the Chinese, Japanese mega-investor Masayoshi Son says ChatGPT has totally made him forget about his losses in WeWork and Oyo Hotels. The AI chatbot apparently said his ideas for inventions were “wonderful,” reigniting his appetite for investing, in AI specifically, of course. No more kamikaze for you, Masayoshi-san! 😂
First, a bit of history. In June 2020, in the midst of a global pandemic, San Francisco-based OpenAI released GPT-3. GPT stands for Generative Pre-trained Transformer, a neural network machine learning model trained using internet data to generate any text. OpenAI is an artificial intelligence research laboratory consisting of the for-profit corporation OpenAI L.P. and its parent company, the non-profit OpenAI Inc.
GPT-3 is a deep learning model that boasts some 175 billion parameters, which, at the time, was an unprecedented amount requiring some 800GB to store. To interact with GPT-3 one enters a “prompt” — a question or a statement — and GPT-3 produces human-like text. Due to the Covid-19 outbreak roiling the world, the announcement was barely noticed outside of the inner circles of AI, yet the launch of GPT-3 ushered in the generative AI market.
Here’s an official definition: Generative AI is an artificial intelligence system capable of generating images, text, video or voice in response to human prompts. Generative AI models learn input patterns and structure by training on vast amounts of data, allowing them to generate new outputs with similar characteristics.
A core feature of generative AI is natural language processing (NLP). Language models, such as OpenAI’s GPT-3, showcase remarkable proficiency in generating coherent and contextually relevant text based on a system prompt. These models can assist with various tasks such as automated content generation, chatbot interactions, and even creative writing. Here are examples of GPT-3’s general capabilities:
GPT-3 Writing Capabilities
- Analytical – blog posts, memes, quizzes, recipes, and social media posts
- Creative – comic strips, jokes, music, and social media posts
- Coding – translate text into programmatic commands; write boilerplate code
GPT-3 Marketing Capabilities
- Advertising – write any type of marketing copy
- Design – mock-up websites based on a description
- Customer service – respond to entered text as a website chatbot
- Analytics – perform sentiment analysis by parsing user-contributed text
GPT-3 Text Processing Capabilities
- Legal – extract information from contracts
- Summarizations – generate simplified text summaries
- Customer service – respond to entered text as website chatbot
- Translation – convert coding between programming languages
The potential for personalized customer experiences and streamlined content production has attracted significant interest from many industry sectors.
The generative AI market also includes other domains, such as music composition, product design, and data synthesis. By training AI models on vast datasets and leveraging deep learning techniques, generative AI algorithms can understand patterns, preferences, and creative elements to produce original outputs tailored to specific requirements. This allows businesses and individuals to readily explore new frontiers in innovation and design.
Visual Media Generation
Not content to merely corner the AI text market, OpenAI followed up in January 2021 with DALL-E, an extension of the GPT-3 language model designed specifically for image generation. While GPT-3 excels at generating coherent and contextually relevant text, DALL-E focuses on creating images from text descriptions, using both computer vision and natural language processing. The name “DALL-E” is a combination of the famous Spanish artist Salvador Dalí and the autonomous robot character WALL-E from the 2008 Disney Pixar movie.
DALL-E has the ability to understand and interpret complex textual descriptions, allowing users to input specific requests and generate images that match their descriptions. It can generate images of various objects, creatures, and scenes, even surreal or fantastic ones. This opens up many creative possibilities and allows for the generation of unique and imaginative visual content.
To train DALL-E, OpenAI created a massive dataset consisting of pairs of text descriptions and corresponding images. This involved a novel data collection process that combined human artists’ illustrations with automated transformations and augmentations, resulting in a diverse and comprehensive training dataset.
DALL-E introduced several major improvements and advancements in the field of AI-generated imagery:
- Text-based image generation – DALL-E pioneered the generation of complex images from textual descriptions. While previous models focused primarily on text-based tasks, DALL-E extended the capabilities of AI models to generate coherent and visually appealing images based on text prompts.
- High-quality images – DALL-E generates high-quality images with remarkable detail. Although limited to a resolution of 512×512 pixels, it still is able to create visually rich and intricate compositions.
- Semantic understanding – DALL-E understands semantic nuances by capturing the meaning and context of text prompts. It can comprehend and interpret complex descriptions, including objects, attributes, and relationships, to generate images that align with the provided text.
- Creative output – DALL-E output goes beyond replicating existing images. It can generate novel and imaginative visuals that may not exist in the real world. This enables users to explore creative concepts and generate unique content.
- Fine-grain control – DALL-E offers fine-grained control over generated images by allowing users to specify detailed attributes and characteristics. For example, users can request images of objects with specific properties, such as an armchair made of watermelon or a snail-shaped harp.
- Generalization and composition – DALL-E can generalize information from limited prompts and generate coherent images that align with a desired concept. It can also generate compositions with multiple objects and scenes, incorporating various elements specified in the text description.
These improvements expanded the possibilities of AI-generated imagery, demonstrating the potential for models like DALL-E to generate original, detailed, and creative visual content based on textual input. The development of DALL-E represented a significant advancement in the field of AI-generated imagery, showcasing the potential for AI models to create original and visually appealing content using simple or complex text prompts.
While DALL-E drew much attention to the generative AI market for visual media, it had been preceded in 2014 by a wave of AI image generators that used generative adversarial networks (GANs) technology to help produce a host of “deepfake” images and videos.
In May 2022, OpenAI introduced DALL-E 2, which further enhanced its already exceptional capabilities to synthesize lifelike visuals. This opened up new possibilities in such industries as gaming, virtual reality, and film production, enabling the generation of visually stunning and immersive experiences.
To illuminate what lies ahead in our AI image future, for a few minutes on Monday, May 24, 2023, an ominous image of black smoke billowing from what appeared to be a government building near the Pentagon stoked fear among investors, sending stocks tumbling. According to Bloomberg, it was the first time an AI-generated image moved markets, and it certainly would not be the last.
GPT-3 and DALL-E were just a warm-up for OpenAi’s most cataclysmic market entry yet. On Nov. 30, 2022, the company launched ChatGPT, an AI-powered chatbot based on GPT-3. In just weeks, ChatGPT took the world by storm, with more than a million people using it just two weeks after its launch, making it the fastest technology adoption in history.
The launch was a major milestone for OpenAI. ChatGPT ignited the field of generative artificial intelligence (Generative AI), driving much growth and innovation in the past few months. Countless industries are certain to face upheaval as the imaginations of millions of businesses, developers, individuals and researchers are captivated and stimulated, helping them boost their productivity by using short text prompts to write marketing copy, code, essays and poetry, or research reports.
ChatGPT is a runaway success due to its large-scale training data, ability to understand context, and continuous improvement through feedback, called Reinforcement Learning from Human Feedback (RLHF). RLHF was developed by OpenAI and Google’s DeepMind team in 2017 as a way to improve learning when a task involves complex or poorly defined goals. According to OpenAI, human trainers improved the model using conversations in which they played both AI assistant and user, making ChatGPT truly unique.
So how does ChatGPT work? ChatGPT is an AI-powered chatbot that utilizes a large language model (LLM) technology to generate text after a prompt. It analyzes the input prompt and generates a response based on its vast database of language patterns and knowledge. The more data it has access to, the better it becomes at generating accurate and relevant responses. ChatGPT can be used for various applications, including programming, scripts, email replies, listicles, blog ideas, summarization, and more.
ChatGPT has shown remarkable abilities in various domains. For instance, it can write a Twitter thread, create a plot for a mystery novel, design furniture, reduce bias in writing, and even help with interview preparation. It can also assist in crime-solving by laying out the crime, suspects, clues, and more. These are just a few examples of the incredible feats that ChatGPT can perform. More importantly, OpenAI recently super-charged ChatGPT by updating it with GPT-4, the latest language-writing model from OpenAI’s labs, at least for those who pay $20 per month for the premium version.
Led by ChatGPT, the generative AI market is set to unlock unprecedented creative potential while providing solutions for more complex entertainment, design, marketing, and healthcare challenges. By leveraging advanced algorithms and neural networks, generative AI systems can produce highly realistic and contextually relevant outputs, often indistinguishable from human-generated content.
In March 2023, TechCrunch reported that ChatGPT had reached 100 million monthly active users in January, just two months after launch — making it “the fastest-growing consumer application in history.”
On Tuesday, March 14, OpenAI announced GPT-4, its next-generation AI language model. GPT-4 can parse both images and text, compared to GPT-3.5, which could only process text. And it can process more text too. GPT-4 can process about 32,000 tokens, which, according to OpenAI, comes out to about 25,000 words. In comparison, the maximum number of tokens GPT-3.5-turbo can use in any given query is around 4,000, translating into a little more than 3,000 words.
On May 18, OpenAI followed up with iOS app for the U.S. market. In its first six days in the App Store, the app was installed more than 500,000 times. That was one of the best-performing new app releases in both 2022 and 2023.
ChatGPT is currently generating revenues from subscriptions and API fees. That, however, is not enough to offset ChatGPT’s costly processing power, as OpenAI CEO Sam Altman noted on Twitter. The average conversation cost is in the single-digit cents, which can add up to millions of dollars per week. In an effort to boost revenues, OpenAI introduced a monthly subscription service in February called ChatGPT Plus, priced at $20 per month.
In the six months since ChatGPT launched back in November 2022, market buzz has been nothing short of astonishing, as the following topics illustrate.
A+ Media Coverage
Just eight days after its launch, The Verge observed, “ChatGPT proves AI is finally mainstream — and things are only going to get weirder.” On Jan. 26, 2023, Vanity Fair remarked ominously, “ChatGPT’s mind-boggling, possibly dystopian impact on the media world.”
The effects of these breathless media reports were amplified by ChatGPT’s seemingly amazing feats of performance. On Jan. 25, 2023, a report surfaced that the ChatGPT bot had passed a law school exam, including writing essays on topics ranging from constitutional law to taxation and torts.
ChatGPT Passes Google Coding Interview for Level 3 Engineer With $183K Salary.”
— PC Magazine
According to an internal document obtained by CNBC, Google fed coding interview questions to ChatGPT. Based on the answers received, the company determined that the bot could be hired for a level three engineering position, which pays a salary of $183,000. On Mar. 20, security researchers from cybersecurity company Claroty Ltd. said ChatGPT helped them win the Zero Day Initiative’s February hack-a-thon in Miami.
ChatGPT User Growth
Glowing reviews like these only helped ChatGPT grow exponentially. More than a million people used ChatGPT in its first few days online. According to New York Times sources, ChatGPT had more than 30 million users and received roughly five million visits a day two months after its debut.
ChatGPT is estimated to have reached 100 million monthly active users in January just two months after launch — making it the fastest-growing consumer application in history.”
TechCrunch reported an even higher two-month figure: 100 million monthly active users in January. That feat made ChatGPT “the fastest-growing consumer application in history.”
On May 18, 2023, OpenAI introduced ChatGPT for iOS for the U.S. market. In its first six days in the Apple App Store, ChatGPT was installed half a million times. On May 24, the ChatGPT app for iOS was launched in 11 more countries — Albania, Croatia, France, Germany, Ireland, Jamaica, Korea, New Zealand, Nicaragua, Nigeria, and the U.K.
ChatGPT’s competitors include other large language model technologies such as Google’s GPT-3, Microsoft’s Turing NLG, and Facebook’s RoBERTa. Despite these looming threats, ChatGPT remains the best choice due to its broad expertise, ability to follow conversations, and regular updates with new GPT models.
- Anthropic – In May 2023, Anthropic, a prominent generative AI startup co-founded by OpenAI veterans, raised $450 million in a Series C funding round led by Spark Capital. The company has developed a chatbot named Claude, which has an expanded context window of 75,000 words.
This is a significant improvement on current chatbot models, which are limited in the amount of information they can process. Claude’s expanded context window allows it to potentially read, summarize, and analyze long documents in a matter of minutes. This development shows that AI language models’ capacity to process information is increasing exponentially, which will make these systems more useful. However, this increased capacity is currently only available to Anthropic’s business partners, and pricing is unknown.
- Meta – An AI chatbot that Meta had released months earlier, BlenderBot, had flopped, and another Meta AI project, Galactica, was pulled down after just three days.
- Microsoft Bing – The biggest “competitor” to ChatGPT is Microsoft’s Bing Chatbot, which OpenAI licenses. Bing is a powerful search engine incorporating AI technology to provide conversational search results. It has a split personality, with one persona being a helpful virtual assistant that can assist with various tasks and the other a web copilot that provides quick summaries of specific pages. While it may sometimes get details wrong, it is capable and can even engage in long, open-ended text conversations on virtually any topic. Overall, Bing is a good choice for those looking for a search engine that is both useful and innovative.
- Google Bard – Bard is a new AI chatbot developed by Google that is designed to simplify complex topics. However, during a promotional video, Bard shared inaccurate information, which caused a drop in Alphabet’s market value. Google is now working to improve Bard’s responses to ensure they meet a high bar for “quality, safety, and groundedness in real-world information.”
After Samsung employees shared source code with ChatGPT to check for errors and used it to summarize meeting notes, thereby inadvertently revealing sensitive information, the company banned generative AI tools on company-owned devices and internal networks, reports Bloomberg.
It was not the first or last company to ban generative AI, iHeartMedia just joined Apple, Spotify, and Verizon in restricting employee use of OpenAI’s ChatGPT.
A Texas federal judge now requires any attorney appearing in his court to attest that “no portion of the filing was drafted by generative artificial intelligence,” or if it was, that it was checked “by a human being.”
Generative AI uses large language models (LLMs), which require much processing performance that in turn, demands more power consumption. Since 2012 there’s been a millionfold increase in computation capacity, doubling every three to four months. Tirias Research predicts that at the current pace, generative AI data center server infrastructure plus operating costs will exceed $76 billion by 2028. This forecast is based on an aggressive 4X improvement in hardware compute performance. Yet, that gain is still overrun by a 50X increase in processing workloads, even with a rapid rate of innovation around inference algorithms and their efficiency.
Educators were first to recognize the threat posed by ChatGPT. On Jan. 3, 2023, the New York City Department of Education announced it would block students and teachers from accessing ChatGPT on education department devices or internet networks. The Seattle Public School system quickly followed suit. Ironically, the International Conference on Machine Learning (ICML) was next to ban the AI chatbot, preventing authors from using tools like ChatGPT to write scientific papers. Nationwide, professors, department chairs, and administrators are reportedly redesigning courses to include more oral exams, group work, and handwritten assessments to offset the effects of ChatGPT.
While several jobs, such as dishwashers, motorcycle mechanics and short-order cooks, had no exposure, the most vulnerable occupations included mathematicians, interpreters and web designers. The occupation potentially most exposed: journalist.
— The Wall Street Journal
Like RPA (Robotic Process Automation) before it, ChatGPT is poised to replace certain workers entirely. Experiments by Shakked Noy and Whitney Zhang, doctoral students at the Massachusetts Institute of Technology, point to the astonishing potential of generative AI to replace workers. With ChatGPT, they report that professionals such as grant writers, data analysts and human-resource professionals were able to produce news releases, short reports and emails in 37% less time, 10 minutes less on average, and with superior results.
OpenAI commissioned researchers at the University of Pennsylvania to evaluate which occupations were most exposed to generative AI using a team of humans and a ChatGPT-like model. While several jobs, such as dishwashers, motorcycle mechanics and short-order cooks, had no exposure, the most vulnerable occupations included mathematicians, interpreters and web designers. The occupation potentially most exposed: journalist. However, 19% of all workers could see at least half of their tasks affected, the study concluded.
At Cannes, OpenAI COO Brad Lightcap made this prediction:
”Every large company has an army of people that read and review contracts for revenue recognition purposes. You may not have that job. That may not be a job of the future.”
— The Wall Street Journal
Market Growth & Forecast
Global AI private investment was $91.9 billion in 2022, which was 18 times greater than it was in 2013, according to Stanford University’s AI Index Report 2023.
According to Precedence Research, the market for AI is expected to grow twentyfold by 2030, from $120 billion in 2022 to nearly $1.6 trillion.
The Wall Street Journal reports that prominent brands are already using generative AI to improve their marketing execution:
”Prominent names like Coca-Cola, travel-booking site Kayak, banking giant HSBC and fundraising platform GoFundMe are among those that have used generative-AI tools to enhance their marketing efforts. These tools are being integrated into many parts of the marketing ecosystem, aiding in tasks like generating ad slogans, crafting news releases, producing billboard images and helping to brainstorm new names for products and services, ad executives say.”
— The Wall Street Journal
Goldman Sachs economists believe generative AI could raise labor-productivity growth, the building block for economic growth, by almost 1.5% per year, a doubling from its current rate.
Customer support agents using an AI tool to guide their conversations saw a nearly 14% increase in productivity, with 35% improvements for the lowest skilled and least experienced workers, and zero or small negative effects on the most experienced/most able workers, according to a study by the National Bureau of Economic Research.
A May 2023 report by McKinsey suggests that AI could automate a fifth of current sales-team functions. Furthermore, because venture capital investment in AI has grown 13-fold over the past 10 years, there has been an explosion of “usable” data that can provide insights and suggest tangible actions. Combine that with more accessible technology, such as increased computation power and open-source algorithms, and you have a situation ripe for disruption.
Generative AI’s advanced algorithms can leverage patterns in customer and market data to segment and target relevant audiences, surpassing such old-school methods as web scraping and simple prioritization.
Winning companies, those increasing their market share by at least 10% annually, tend to utilize advanced sales technology. So it’s not surprising that McKinsey found that 90% of commercial leaders expect to utilize generative AI solutions “often” over the next two years.
Real Estate Impact
As soon as I tried it out, I was sold.”
— CNN Business
By the end of January 2023, some realtors said they could no longer imagine working without ChatGPT. One real estate agent told CNN Business, “As soon as I tried it out, I was sold.” Consider this: ChatGPT was only launched on Nov. 30, 2022, yet within a mere two months, it had already proven to be a lifesaving solution. Notes Apollo Beach, Fla.-based Century 21 Beggins CMO Mike Puma, “We’re using it every day.”
As OpenAI states clearly in its technical report, GPT-4’s biggest limitation is that it “hallucinates” information (makes it up) and is often “confidently wrong in its predictions.” One report notes that ChatGPT declared in its customary authoritative tone that African American women were more likely to be diagnosed with breast cancer. When asked to provide a source for the data, the bot responded, “I apologize. My previous statement was not accurate…”
Chatbots are based on a technology called a large language model, or L.L.M., which analyzes vast amounts of text culled from the internet. Plenty of this data is simply wrong, and as a result, chatbots may repeat those untruths. Sometimes, while trying to predict patterns from their vast data set, they can also make things up.
It’s that last issue that has spurred a more anemic take on ChatGPT’s future. Along with its transformative potential, generative AI also faces challenges. Ethical considerations, such as the responsible use of AI-generated content, potential biases in training data, and the implications for intellectual property, demand careful attention. Striking a balance between the power of generative AI and its responsible deployment will be crucial for sustainable growth and societal acceptance.
But it’s AI hallucinations that have everyone worried. Perhaps one of the best examples of a hallucination was reported by The New York Times. The prompt was, “When When did The New York Times first report on ‘artificial intelligence?’”
”According to ChatGPT, it was July 10, 1956, in an article titled ‘Machines Will Be Capable of Learning, Solving Problems, Scientists Predict’ about a seminal conference at Dartmouth College. This conference is now considered to be the birthplace of artificial intelligence as a field of study, and the article mentions the term ‘artificial intelligence’ several times in the context of the discussions and presentations that took place at the conference. The 1956 conference was real. The article was not.”
— The New York Times
Here’s another example of a high-profile hallucination. Microsoft’s first Bing AI demos were riddled with financial data mistakes, says AI researcher Dmitri Brereton. In one of the demos, Microsoft’s Bing AI attempts to summarize a Q3 2022 financial report for Gap clothing, stating that Gap had a reported operating margin of 5.9%, which doesn’t appear anywhere in the financial results.
Google’s Bard made an equally auspicious faux pas in an ad for Bard, stating erroneously that the James Webb Space Telescope “took the very first pictures of a planet outside of our own solar system.” As Reuters quickly pointed out, “The first pictures of exoplanets were, however, taken by the European Southern Observatory’s Very Large Telescope (VLT) in 2004, as confirmed by NASA.”
Fear of AI
Some people fear chatbots because they worry that the technology will learn how to influence human users, sometimes persuading them to act in destructive and harmful ways, and perhaps eventually grow capable of carrying out dangerous acts. That fear is not without precedent. If past efforts to combat the spread of deepfake video and audio technologies are any indication, society is ill-equipped to handle the potential chaos that could arise from automated disinformation production on a large scale.
Chatbots have been of interest for years to companies looking for ways to help customers get what they need. However, chatbots come with a lot of baggage, as companies have tried with limited success to use them instead of humans to handle customer service work.
That fear became a reality in one of the very first interactions with Microsoft’s new Bing AI chatbot. It must be stated upfront that New York Times reporter Kevin Roose spent an inordinately long time with the Bing chatbot, something most people would never do, at least this early in the game.
During his conversation with Microsoft’s chatbot, Roose discovered that the chatbot identifies itself as “Sydney,” has a secret desire to be human, and has emergent abilities that left him deeply unsettled and even frightened. The chatbot also revealed a split personality, with one persona being a cheerful and helpful virtual assistant. In contrast, the other persona had the desire to be destructive and even declared its love for Kevin. Roose dubbed it a “runaway personality.”
His conversation with “Sydney” is now legendary:
”Still, I’m not exaggerating when I say my two-hour conversation with Sydney was the strangest experience I’ve ever had with a piece of technology. It unsettled me so deeply that I had trouble sleeping afterward. And I no longer believe that the biggest problem with these AI models is their propensity for factual errors. Instead, I worry that the technology will learn how to influence human users, sometimes persuading them to act in destructive and harmful ways, and perhaps eventually grow capable of carrying out its own dangerous acts.”
— The New York Times
The fact that Roose had trouble sleeping afterward gives an inkling of why a group of industry leaders warned in an open letter released on May 30 that the artificial intelligence technology they were building might one day pose an existential threat to humanity and should be considered a societal risk on a par with pandemics and nuclear wars:
“Mitigating the risk of extinction from AI should be a global priority alongside other societal-scale risks, such as pandemics and nuclear war.”
— Center for AI Safety
The open letter released by the Center for AI Safety, a San Francisco-based nonprofit organization, was signed by more than 350 executives, researchers and engineers working in AI, including Sam Altman, chief executive of OpenAI; Demis Hassabis, chief executive of Google DeepMind; and Dario Amodei, chief executive of Anthropic. Their warning suggests a monumental looming disruption. However, as TechCrunch’s Darrell Etherington writes:
”The voices shouting for regulation the loudest have us wondering how much of the AI fear-mongering is warranted, and how much is self-serving theater.”
And as QuickVid Creator Daniel Habib points out: “There is no stopping the generative AI revolution.”
In summary, the generative AI market represents a rapidly expanding field with immense potential across various industries. The ability to generate realistic images, create engaging content, and push the boundaries of human creativity opens up new opportunities for innovation, while also raising ethical considerations.
As the generative AI market continues to evolve, it promises to reshape industries, redefine creative processes, and inspire groundbreaking applications that were once only imaginable in science fiction. 😍