ChatGPT Storm: Taking on the Tech Giants
Are Tech Giants safe?
A new term called ChatGPT is making waves in the IT industry. On December 1, the system went live for public beta testing. After only one week of operation, ChatGPT has surpassed 1 million users. There are rumors that the platform will eventually replace Google search and, more importantly, people. Wondering how it works because you are amazed? You’re in the correct spot, I suppose.
The business claims that Reinforcement Learning from Human Feedback was used to train the AI platform (RLHF). Simply put, the human AI trainers delivered dialogues in which they played the roles of the user and the AI assistant. Additionally, the company had provided trainers with access to sample written suggestions, which assisted the trainers in creating their responses.
Additionally, OpenAI developed a reinforcement-learning reward model. This indicates that the business intended to gather some comparative information for more effective response techniques. The company used the chatbot conversations with AT trainers to gather this data tech giants. The business chose a written mode message, sampled a few different completions, and had AT trainers rank the messages. As a result, it is taught how to respond in detail to a prompt and follow instructions. Users only need to enter their question, and the chatbot will respond. But how does it differ from other chatbots that use artificial intelligence (AI)? The developers claim that ChatGPT, unlike other AI chatbots, has the ability to respond to follow-up queries, acknowledge mistakes, refute false premises, and reject unsuitable requests tech giants.
ChatGPT is a free service, according to OpenAI, but only while the research preview is in effect. This indicates that utilizing the platform is as simple as going to the OpenAI website and clicking the “Try ChatGPT” button. To utilize the ChatGPT, you may either create an account or log in with your OpenAI credentials. Additionally, the business has provided a sample for reference on the website.
The newest member of the Generative Pre-Trained Transformer (GPT) family is ChatGPT. It is the most recent AI tool for automatically generating text, to put it simply. However, it is not without flaws or restrictions. On its website, OpenAI acknowledges that ChatGPT occasionally generates plausible-sounding but mistaken or absurd responses. The model also uses specific words and phrases far too frequently.
The age of AI might finally be upon us.
You may be familiar with ChatGPT, a chatbot powered by artificial intelligence that was recently made available to the public by startup OpenAI. The startlingly competent AI chatbot, which was released on Nov. 30, amassed 1 million users in just five days. College instructors are now worried about the emergence of AI-generated College essays in the aftermath of ChatGPT’s release, while software developers may be worried about the rise of ChatGPT’s AI-generated coding skills.
Microsoft (MSFT, 4.69%) clearly sees great potential in this new, cutting-edge AI engine, as it has already invested $1 billion in OpenAI in 2019 and is reportedly in talks to invest an additional $10 billion in the company. Microsoft this week introduced an OpenAI service on its Azure platform, which programmers may now include in their software creations. Microsoft is also aiming to integrate ChatGPT’s capabilities into its current software products, from Office to Bing. Microsoft CEO Satya Nadella stated last week at the World Economic Forum that AI will become “mainstream” in “months, not years.”
One can be confident that every major tech giants firm will now be spending substantially on AI to compete, with the tech giants appearing to have crossed a threshold where it is now considered standard in a variety of commercial and consumer applications tech giants.
If the AI wars are starting, then certain “weapons dealers”—the semiconductor firms that make AI function—should reap significant financial rewards, possibly even more so than Microsoft or other significant tech giants platforms.
As the saying goes, during the California gold rush of the 1800s, the manufacturers of picks and shovels made more money than the prospectors themselves. The following “picks and shovels” semiconductor stocks should prosper greatly in the coming years if AI is applied to them.
Graphics processing units (GPUs) are essential for running artificial intelligence, and Nvidia (NVDA 3.66%) is the market leader in this area. The fact that ChatGPT doesn’t even currently work on Nvidia’s newest processor, the H100, or “Hopper,” which was just announced late last year, is intriguing and appealing. Instead, the current version of ChatGPT uses the two-year-old A100 chips tech giants. If you think ChatGPT is impressive now, get ready. The H100, which began shipping in October, is projected to perform AI “training” functions at nine times the speed of the A100 and “inference,” or the act of an AI reacting to a question or other stimulus, at 30 times the speed. In addition, the H100 promises 3.5 times better energy efficiency and a three times lower total cost of ownership.
This year, Nvidia also introduces Grace, its first CPU, which is created exclusively to cooperate with Hopper and Nvidia’s networking-focused data center processing units (DPUs). Together, these full AI systems should be able to perform ultra-fast GPU processing, ultra-fast networking, and intra-system data movement.
The video game pandemic hangover and bitcoin bear market caused Nvidia’s gaming chip sales to decline in the most recent third quarter, while its data centre revenue increased by an amazing 31%. In 2023, expect Nvidia to gain from the debut of Hopper H100 processors and the beginning of the AI wars. That is especially true following the stock’s 50% collapse in 2022, when the price has since fallen to a more acceptable level.
Taiwan Semiconductor Manufacturing and ASML Holdings
These two foreign firms, Taiwan Semiconductor Manufacturing Corporation (TSM 2.02%) and ASML Holdings (ASML 2.66%), are necessary for the fabrication of Nvidia GPUs. Since the Oracle of Omaha purchased $4 billion worth of TSMC stock last summer, you may recognize it as one of his most recent stock picks.
First, let’s talk about ASML, which holds the monopoly on an essential technique called extreme ultraviolet lithography (EUV) that is used to produce cutting-edge semiconductors. The ultra-fine lasers of EUV tech giants became required as cutting-edge chip makers started producing semiconductors with transistors spaced less than 10 nm apart. Given that the most power-efficient, transistor-dense processors are needed for artificial intelligence applications to function, the escalating AI wars will put pressure on chip foundries to develop an increasing number of cutting-edge chips. That implies more people will buy ASML’s equipment.
All in all, lithography is only one of several procedures involved in making a cutting-edge semiconductor tech giants. To produce these very intricate chips without any flaws, lithography has to collaborate with masking tools, etch and deposition machines, metrology and inspection tools, and improved packaging.
That’s really, really challenging, which is why Taiwan Semiconductor—the biggest outsourced foundry in the world—has developed its own competitive moat over the years. In terms of manufacturing cutting-edge chips, TSMC long ago outpaced Intel (INTC, 3.8%), and given TSMC’s size and scale advantage as well as Intel’s present challenges as a result of its exposure to the sluggish PC market, it only seems to be increasing that lead. In the fourth quarter, TSMC manufactured more than 56% of the semiconductors used in the world, with an even larger percentage used in cutting-edge products.
TSMC management highlighted its high-performance computing sector for AI clients as the cause for its optimism regarding a rebound in the semiconductor business in the second half of 2023 on its most recent results call with investors.
As a result, ASML and TSMC collaborate, but both have a competitive advantage when it comes to making the most sophisticated chips needed for AI processing. Value investors like Buffett might favour TSMC because it is less expensive and trades at only 13.7 times earnings. A greater value of 42 times earnings has been made possible by ASML’s exclusivity in the tech giants sector, asset-light business strategy, and smoother growth prospects tech giants.
Finally, massive amounts of memory and storage are needed to process all that data, which should benefit memory manufacturer Micron Tech giants (MU 0.91%).
Because of the unprecedented decline in PC sales as well as the difficulties in cellphones and consumer electronics, the delicate supply-demand balance in the memory market has been overloaded, and Micron’s results are currently in freefall. Its earnings will likely be negative for the next two quarters.
Even though Micron’s performance has drastically declined, the stock has miraculously managed to hold steady at levels similar to those of last June. This is because the market is typically forward-looking and the stock now trades slightly above Micron’s book value. Micron is one of only three companies worldwide that manufactures DRAM memory at scale and is one of only five companies that manufactures NAND flash storage.
Due to the lack of rivalry, Micron and peer SK Hynix have announced significant budget cutbacks for 2023, which should aid in reestablishing the supply-demand balance in the second half of the year. Analysts now doubt that market leader Samsung will maintain its investment in memory chips despite the downturn, particularly in light of its worse-than-expected fourth quarter earnings preannouncement. In fact, Digitimes reported on January 16 that Samsung would reduce part of its NAND output, maybe recognising the reality of the situation.
Additionally, last year, Micron outperformed this small competition in terms of tech giants. Micron became the first memory manufacturer to produce 232-layer NAND flash chips and 1-beta DRAM chips in just the last six months. Micron should profit in the second half of 2023 and beyond as demand for memory-intensive AI servers picks up steam due to its current technological advantage and the fact that all market players currently appear to be reducing production.
Author Ayesha Urooj holds master’s degrees in both research writing and computer science. The author pays close attention to details and is committed to giving readers engaging information.
About Remote IT Professionals
Remote IT Professionals is devoted to helping remote IT professionals improve their working conditions and career prospects.
We are a virtual company that specializes in remote IT solutions. Our clients are small businesses, mid-sized businesses, and large organizations. We have the resources to help you succeed. Contact us for your IT needs. We are at your service 24/7.