{"id":60462,"date":"2023-04-20T04:45:10","date_gmt":"2023-04-20T08:45:10","guid":{"rendered":"https:\/\/www.technologyinvestor.com\/?p=60462"},"modified":"2023-04-20T04:45:10","modified_gmt":"2023-04-20T08:45:10","slug":"chaos-abounds-but-so-do-opportunities-here-are-the-stocks-to-benefit-from-the-new-ai","status":"publish","type":"post","link":"https:\/\/www.technologyinvestor.com\/?p=60462","title":{"rendered":"Chaos abounds. But so do opportunities. Here are the stocks to benefit from the new AI"},"content":{"rendered":"

View this email in your browser. Click here<\/a><\/strong>.<\/p>\n

There’s blood in the water, specifically:<\/p>\n

+ Commercial real estate.<\/p>\n

+ New companies that can’t get venture financing and can’t go IPO.<\/p>\n

“Distress” deals are emerging. The key for investors like us is to be super careful. Reject one great deal today. A better one will come along tomorrow. It’s the old adage: When God closes a door; He opens a window. And usually very fast.<\/p>\n

Keep your powder dry with three month treasuries — which are gloriously paying over 5%.<\/p>\n

Our progress as a society has always been technology. From electricity to cell phones. From fiber to the Internet. (And a zillion others.) Each invention has brought great benefits and changed the way we live. But all were adopted slowly. It took years to bring electricity to the factory. And years for the Internet to change our shopping habits.<\/p>\n

The biggest and fastest is now upon us. It’s called generative Artificial Intelligence. Do not dismiss this as Harry’s Ravings.\u00a0 You will remember that on February 9, I\u00a0 wrote “The most important blog I have ever written<\/em>” This was my headline:<\/p>\n

Generative AI \u2014 e.g. ChatGPT \u2014 is the most powerful invention I have ever seen. Here\u2019s why. This is the most important blog I have ever written.<\/strong><\/p>\n

That blog is here.<\/strong><\/a><\/p>\n

Since then every major news outlet has written and broadcast about the new AI,\u00a0 including 60 Minutes<\/em> last Sunday. (Click here<\/strong><\/a>)<\/p>\n

In recent days, I have read perhaps 30 articles and research reports on the subject. I can’t list them all. A new one from the Economist<\/em>\u00a0concludes:<\/p>\n

The capital flowing into generative-ai startups, which last year collectively raised $2.7bn in 110 deals, suggests that venture capitalists are betting that not all the value will be captured by big tech. Alphabet, Microsoft, their fellow technology titans and the Chinese Communist Party will all try to prove these investors wrong. The ai race is only just getting started.<\/p>\n

For the Economist piece, click\u00a0here.<\/strong><\/a><\/p>\n

The New York Times ran a piece “35 Ways Real People Are Using A.I. Right Now”<\/strong><\/p>\n

The public release of ChatGPT last fall kicked off a wave of interest in artificial intelligence. A.I. models have since snaked their way into many people\u2019s everyday lives. Despite\u00a0their\u00a0flaws, ChatGPT and other A.I. tools are helping people to save time at work, to code without knowing how to code, to make daily life easier or just to have fun.<\/p>\n

It goes beyond everyday fiddling: In the last few years, companies and scholars have started to use A.I. to supercharge work they could never have imagined, designing new molecules with the help of an algorithm or building alien-like spaceship parts.<\/p>\n

Here\u2019s how 35 real people are using A.I. for work, life, play and procrastination.<\/p>\n

For the NYTimes’ piece, click here.<\/strong><\/a><\/p>\n

I’m awed by it all. I asked a very talented researcher\/reporter, Richard Grigonis, to help me out, Lay out the industries it will affect. And detail the companies who will benefit and which you and I could invest in. Here’s his brilliant report:<\/p>\n

Where Generative AI Stands Now <\/strong>by Richard Grigonis<\/p>\n

Things are happening so rapidly in the generative AI field that it\u2019s almost impossible to keep up.<\/p>\n

Lots of people are worried that generative AI (ChatGPT-4, etc) will put them on the unemployment lines. There is renewed chatter about the need for guaranteed \u201cbasic income\u201d and all that.<\/p>\n

There\u2019s an insightful blog post on AI\u2019s possible impact on jobs at Packy McCormick\u2019s \u201cThe Not Boring\u201d blog. The article uses two economic concepts to argue that 1) the availability of AI will unleash a \u201clatent\u201d demand for services, while 2) increasing capabilities will result in people expecting faster and higher-quality services. AI-driven increased demand implies that job displacement is less likely, as the need for output will surge. Designers will be expected to produce better and faster work; the ease of creating legal documents will lead to better docs (and perhaps more lawsuits; and consultants will need to offer more efficient advice).<\/p>\n

It reminds me of when the U.S. superhighway system was built: It encouraged more people to drive, and they drove farther distances to go places they had never gone before and do things they had never done before. It brought about expansion of long-distance trucking.<\/p>\n

The many industries and activities most likely to benefit from generative AI include the following:<\/p>\n

+ Software and Website coding<\/strong>: Programmers spend a lot of time writing code, and they don\u2019t actually produce many lines of code per day, since it all has to be tested and debugged. GitHub Copilot, the result of collaboration between the GitHub code development environment and OpenAI, assists in the production and debugging of code. In one study, the GitHub founders discovered that developers using GitHub Copilot coded up to 55% faster than those who didn\u2019t. And NVIDIA\u2019s Jensen Huang as said that, with generative AI, anyone \u2013 be they brilliant, fair-to-middling or dumber than dirt \u2013 can now program a computer using natural language. As for websites, OpenAI founders have demoed the ability of AI to read a handwritten note describing a website and then build a whole working site from it.<\/p>\n

+ Drug Design:<\/strong> Generative AI can accelerate the drug discovery and development process by pinpointing possible drug candidates and evaluating them through computer simulations before advancing to clinical trials involving animals and humans. It normally takes three to six years and costs an average of $1.8 billion to bring a drug to market. By employing generative AI, pharmaceutical companies can design drugs within months at far less cost. Indeed. Gartner says that of the more than $1.7 billion invested by venture capital firms in generative AI solutions over the last three years, AI-enabled drug discovery and AI software coding received the most funding.<\/p>\n

+ Materials Science:<\/strong> In the past materials with desired properties came about by trial-and-error \u2014 think of rubber, which melted in the summer and cracked in the winter until Charles Goodyear accidentally dropped some sulfur into it on a hot stove, creating the first vulcanized rubber. Now, using what\u2019s called \u201cinverse design\u201d AI helps engineers create new materials with targeted properties by building molecular computer models atom-by-atom. In this area, generative AI revolutionizes a whole slew of sectors such as automotive, aerospace, defense, energy, electronics and medical.<\/p>\n

+ Performance and Manufacturing Methods:<\/strong> Device parts can be optimized for specific goals and constraints. Automakers, for example, can use AI to develop lighter designs for fuel efficiency.<\/p>\n

+ Chip Design:<\/strong> Generative AI can optimize component placement in semiconductor chip design using machine learning techniques. It shrinks the product-development life cycle from weeks to hours.<\/p>\n

+ Synthetic Data:<\/strong> Synthetic data (also called data fabrication) is data that\u2019s artificially generated by AI rather than gathered from real-world observations. This gets around problems of ensuring privacy in the case of, say, patient data used for medical research and analysis.<\/p>\n

+ Advertising:<\/strong> Generative AI can help (or dominate) with copywriting, storyboarding, writing jingles for commercials, and advertisements based on improved versions of existing ones.<\/p>\n

+ Art and Design:<\/strong> Generative AI helps artists and designers (and anybody else) create new works by effortlessly generating images based on new ideas and concepts.<\/p>\n

+ News Media:<\/strong> Generative AI can summarize research and write and help produce news stories and web content. There\u2019s still a need for editors and fact checkers, though. Maybe one day a \u201ctalking heads\u201d pundit show will consist of a bunch of AI-written opinion pieces read by AI-generated talking heads, who knows?<\/p>\n

+ Entertainment:<\/strong> Generative AI can assist in the creation of video games, movies, and TV shows, so content creators can be more prolific and don\u2019t need a big staff.<\/p>\n

+ The Video Future:<\/strong> Although it\u2019s not quite there yet, rapid progress is being made on text-to-video. So not only will many folks at places like Pixar be out of a job, but in a few years you\u2019ll be able to write a screenplay and have a movie generated, with whatever actors you want, living or dead \u2014 and even insert yourself in the movie. Of course, you could just tell GPT-10 (or whatever exists at that time) to write the screenplay for you, after you have a brief conversation with it to describe the scene, characters and plot. The resulting movie will be viewable on virtual reality goggles in 3D. And then, several years beyond that, something akin to the holodeck from Star Trek will become reality.<\/p>\n

Richard Grigonis’ Stock Picks:<\/strong><\/p>\n

+ NVIDIA (NVDA)<\/strong> has shot up more than 83% since October because NVIDIA\u2019s GPU technology is the only real game in town for training and inference processing ChatGPT and its fellow generative AI models. To train the new, even bigger GPT-5 model, the latest and greatest generation of NVIDIA equipment is necessary, and tons of it must be manufactured, tested and optimized. OpenAI and Microsoft are obviously waiting for this to happen.<\/p>\n

Additionally, NVIDIA’s has stepped forward with its own cloud services offering, AI Foundations, for smaller businesses unable to develop their own AI models from scratch. With it they can “build, refine and operate custom large language models and generative AI models that are trained with their own proprietary data and created for their unique domain-specific tasks.”<\/p>\n

The models include NeMo, NVIDIA\u2019s language models; BioNemo, a drug and molecule discovery version of the NeMo model built for the medical research community; and Picasso, capable of generating images, video and \u201c3D applications\u2026 to supercharge productivity for creativity, design and digital simulation,\u201d according to NVIDIA\u2019s March release. NeMo and Picasso both run on NVIDIA’s new DGX Cloud platform and companies will be able to access them via an online portal.<\/p>\n

+ Super Micro Computer Inc (SMCI),<\/strong> dba Supermiro, is the other major player in the generative AI area. Those NVIDIA GPU\u2019s have to plug into somebody\u2019s computer platform, and Supermicro announced on March 21st that it was now shipping its top-of-the-line new GPU servers featuring the latest NVIDIA HGX H100 8-GPU system. Note: it takes a whole NVIDIA 8-GPU node to run a user GPT-4 query on OpenAI and Microsoft\u2019s cloud servers, mostly because of the huge memory requirements to hold the large AI model in the HBM memory of NVIDIA A100 or H100 Tensor Core GPUs.<\/p>\n

+ Microsoft Corp. (MSFT)<\/strong> made perhaps the wisest investment in its history when it announced a multiyear, multibillion dollar investment in OpenAI in January 2023. They are using OpeAI\u2019s GPT technology to revamp the Bing search engine and bring AI to their other products running in the enterprise.<\/p>\n

+ Alphabet Inc. (GOOG, GOOGL)<\/strong> is playing catchup to OpenAI and Microsoft. Ironically, OpenAI didn\u2019t invent the natural language model underlying its famous ChatGPT. GPT stands for \u201cgenerative pretrained transformer,\u201d and it was Google that invented the transformer language model in 2018 with BERT (Bidirectional Encoder Representations from Transformers). BERT is now used to enhance Google\u2019s search and translation functions. Still, Google has high hopes for its Bard chatbot, based on its LaMDA language model, claimed to be more powerful than ChatGPT. Google is also readying its Imagen AI image generator that\u2019s said to be at least as good as OpenAI\u2019s DALL-E2. Moreover, the Alphabet subsidiary DeepMind also has a powerful chatbot called Sparrow that may be released soon.<\/p>\n

+ Meta Platforms Inc. (META)<\/strong> is better known for social media (Facebook, Instagram, etc.) and online advertising. Meta is investing heavily in AI and scrambling to infuse the technology into its products. It\u2019s even building an AI supercomputer, presumably to train its own AI language models.<\/p>\n

+ Taiwan Semiconductor Manufacturing Co. Ltd. (TSM)<\/strong> is the largest pure-play semiconductor foundry but is only now trying to take advantage of the generative AI boom, so there\u2019s some room for growth. The only wild card here might be if mainland China attacks Taiwan.<\/p>\n

+ Baidu Inc. (BIDU)<\/strong> the Chinese internet search provider, launched its own enterprise large language model called Wenxin Qianfan in March. At the moment they seem to be a \u201cme-too\u201d company in this area, rushing to make a name for themselves.<\/p>\n

+ Arista Networks Inc. (ANET)<\/strong> offers high-throughput data center switches sporting bandwidth able to interconnect the processing power needed for training and using AI models.<\/p>\n

+ Relx PLC (RELX)<\/strong> is using AI to revolutionize the way it provides its professional and business customers with information and analytics. The magic \u201cAI aura\u201d seems to have enhanced its revenue growth, as it has many companies.<\/p>\n

Here’s how ChatGPT organized Harry’s Switzerland trip<\/strong><\/span><\/p>\n

In late May, Susan is going with the girls to Sicily. She’s leaving me all by myself to play tennis and to take care of Rosie, our 12-year-old Javanese. What a great opportunity to indulge my love for mountains (which Susan doesn’t like.) Why not ask AI to make me a trip?<\/p>\n

My question to ChatGPT:<\/strong><\/p>\n

I‘d like to plan a seven day trip around the mountains of Switzerland in late May 2023. I don’t want to rent a car. All my travel should be by train, bus and taxi.\u00a0<\/em><\/p>\n

ChatGPT\u2019s answer:<\/strong><\/p>\n

A seven-day trip around the mountains of Switzerland in late May 2023 can be a wonderful experience. Switzerland is well-connected with an extensive public transportation network that includes trains, buses, and taxis. Here is a suggested itinerary for your trip:<\/p>\n

Day 1: Zurich to Interlaken Take a train from Zurich to Interlaken, a beautiful town located between Lake Thun and Lake Brienz. Spend the day exploring the town and taking in the stunning views of the surrounding mountains.<\/p>\n

Day 2: Interlaken to Grindelwald Take a bus or a train from Interlaken to Grindelwald, a popular destination for outdoor activities and skiing. Spend the day hiking or skiing in the mountains, and take a cable car up to the top of the First Mountain for breathtaking views.<\/p>\n

Day 3: Grindelwald to Zermatt Take a train from Grindelwald to Zermatt, a picturesque mountain town famous for its view of the Matterhorn. Explore the town and take a cable car up to the Klein Matterhorn for stunning views of the surrounding mountains.<\/p>\n

Day 4: Zermatt to St. Moritz Take a train from Zermatt to St. Moritz, a glamorous resort town known for its skiing and luxury accommodations. Spend the day exploring the town and taking in the beautiful views of the surrounding mountains and Lake St. Moritz.<\/p>\n

Day 5: St. Moritz to Davos Take a train from St. Moritz to Davos, a charming town known for its skiing and winter sports. Spend the day exploring the town and taking in the beautiful views of the surrounding mountains.<\/p>\n

Day 6: Davos to Lucerne Take a train from Davos to Lucerne, a historic town located on Lake Lucerne. Spend the day exploring the town and taking in the beautiful views of the lake and surrounding mountains.<\/p>\n

Day 7: Lucerne to Zurich Take a train from Lucerne to Zurich, where you can spend the day exploring the city and taking in its rich history and culture.<\/p>\n

Throughout your trip, you can use taxis or public transportation to get around each town and between destinations. Swiss public transportation is efficient, reliable, and easy to navigate, so you should have no trouble getting around during your trip.<\/p>\n

My next question to ChatGPT:<\/strong><\/p>\n

Which hotels would you suggest for my seven day trip to Switzerland?<\/em><\/p>\n

ChatGPT\u2019s answer:<\/strong><\/p>\n

There are many excellent hotels to choose from in Switzerland, depending on your budget, preferences, and travel itinerary. Here are some suggestions for hotels in the towns you will be visiting during your seven-day trip:<\/p>\n

    \n
  1. Interlaken: Hotel Victoria-Jungfrau Grand Hotel & Spa This luxurious hotel is located in the heart of Interlaken and offers stunning views of the surrounding mountains. It features an award-winning spa, gourmet restaurants, and elegantly appointed rooms and suites.<\/li>\n
  2. Grindelwald: Romantik Hotel Schweizerhof This charming hotel is located in the center of Grindelwald and offers traditional Swiss hospitality and cozy accommodations. It features a spa, restaurant, and bar, and is just a short walk from the train station.<\/li>\n
  3. Zermatt: The Omnia This contemporary hotel is located in the heart of Zermatt and offers spectacular views of the Matterhorn. It features a spa, restaurant, and rooftop terrace, as well as modern and stylish rooms and suites.<\/li>\n
  4. Moritz: Badrutt’s Palace Hotel This iconic hotel is located in the heart of St. Moritz and is a symbol of Swiss luxury and elegance. It features a spa, multiple restaurants, and bars, as well as opulent rooms and suites with mountain views.<\/li>\n
  5. Davos: Steigenberger Grandhotel Belv\u00e9d\u00e8re This grand hotel is located in the center of Davos and offers classic Swiss charm and hospitality. It features a spa, restaurant, and bar, as well as spacious and comfortable rooms and suites.<\/li>\n
  6. Lucerne: Palace Luzern This elegant hotel is located on the shores of Lake Lucerne and offers stunning views of the lake and mountains. It features a spa, restaurant, and bar, as well as luxurious rooms and suites with modern amenities.<\/li>\n<\/ol>\n

    These hotels are just a few examples of the many excellent accommodations available in Switzerland. Be sure to book in advance as they can get fully booked during the high season.<\/p>\n

    All this took ChatGPT less than three minutes.<\/p>\n

    This week’s New Yorker<\/em>\u00a0had the final word:<\/p>\n

    \"\"<\/p>\n

    Silly, but fun.<\/strong><\/span><\/p>\n

    \"\"<\/p>\n

    \"\"<\/p>\n

    That’s it for today. I’m off to see two grandchildren in Maine and celebrate the opening of the new Barry Mills building at Bowdoin College which my daughter attended and at which she met her special Ted, now my son-in-law.<\/p>\n

    It’s Spring. Our daffodils are out. What a wonderful time to be alive.<\/p>\n

    \"\"<\/p>\n

    — Harry Newton<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

    View this email in your browser. Click here. There’s blood in the water, specifically: + Commercial real estate. + New companies that can’t get venture financing and can’t go IPO. “Distress” deals are emerging. The key for investors like us is to be super careful. Reject one great deal today. A better one will come […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-60462","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"_links":{"self":[{"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/posts\/60462","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=60462"}],"version-history":[{"count":34,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/posts\/60462\/revisions"}],"predecessor-version":[{"id":60506,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=\/wp\/v2\/posts\/60462\/revisions\/60506"}],"wp:attachment":[{"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=60462"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=60462"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.technologyinvestor.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=60462"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}