Uncover Tech Insights - November 13-17
Weekly market stock moves
Former Group head at Microsoft who worked on Bing shares valuable insights
Some thoughts from the interview, together with my thinking on the Google Safari Default topic:
Building a successful Search Engine is not as easy as people think. The main thing is having a ton of data and, with it signal. Over the years, Google has built itself into a position where it has so much data flowing through it daily that it makes its product superior.
Defaults are very powerful, and Apple has multiple times thought about having its own search engine. However, it's clear that initially, it wouldn't be as good as Google Search because it lacks the data to be that good. It could take 6 months or even more to reach a similar level, if at all, and in the meantime, what would users do? Would they really wait 6+ months or start switching over?
For Apple, it's much smarter and more predictable to get the 36% revenue cut from Google and avoid the costs of development and business success risk.
For Google, the deal was great at the start because they saw how quickly mobile was growing and wanted to secure their part in the next computing platform. It probably still makes sense for them because they don't want any competitor to enter Search, especially not one that owns the hardware platform. At the same time, they continue to get signal and data at the biggest possible scale, which drives their whole flywheel.
Former Global Head at Google shares his view on the company and its challenges with the new AI wave
The expert thinks OpenAI is a threat to Google, although he thinks OpenAI will go more into the enterprise road as it makes more business sense for them. The expert thinks the real risk for Google could be from a new company leveraging the tech of OpenAI and redoing the Search experience.
The new AI products will help a lot of small advertisers because of better information and assisting with the campaign's creative. “That means that not just the Fortune 100, but now the next 20,000 advertisers will have the ability to make really interesting good decisions."
The expert still sees an uncertain short-term environment for advertising and has brought up the potential of “squeezing” big advertisers and pulling forward some demand to meet short-term targets for the last quarter as a possible scenario.
According to the experts, a lot comes down to whether the new AI tech is revolutionary or just an incremental change. If it is an incremental change, then Google is really well positioned for it, if it is revolutionary, then Google will have more problems as it faces the classic Innovator’s dilemma.
The expert is not worried about higher costs related to AI because AI is more compute-heavy than, for example, Google Search query business. Expert thinks Google is very good when it comes to driving down costs and increasing efficiency in data centers over time.
Former Senior Manager at PayPal responsible for managing B2B payments shares his insights into the company and the industry
PayPal did a lot of M&A, but according to the expert, the main goal was to acquire customers. Integrations of these products would often take years to complete as many of these platforms are not built on the same stack as PayPal.
Venmo was kept separate by choice from PayPal; however, the backend got integrated.
According to the expert, payment processing is a low-margin business; however, your IT spending and fraud losses define the overall margin. PayPal has the advantage of being a two-sided network with data on customers and merchants. That can affect fraud detection.
According to the expert, PayPal was a bit slow with an API integration and ease of use but has since caught up to their competitors.
The expert thinks the EU is too challenging and that PayPal should focus on the US and Latin America.
The expert also thinks the two-sided marketplace will help PayPal regarding AI development.
Most impactful news of the week
OpenAI's board made a surprising revelation by announcing the removal of CEO Sam Altman, citing concerns about his lack of consistent transparency in communications with the directors. In the aftermath, Mira Murati, the company's Chief Technology Officer, was appointed as the interim CEO. Shortly thereafter, OpenAI President Greg Brockman, responsible for overseeing crucial products like ChatGPT, resigned in response to Altman's removal. The unexpected departures took even the company's staunchest supporters by surprise. High-ranking executives at Microsoft, a major investor that had committed over $10 billion to OpenAI earlier in the year, were only informed of the board's decision moments before the public announcement.
OpenAI expresses optimism about the potential return of key departed figures, including Sam Altman and Greg Brockman, following Altman's abrupt dismissal on Friday. Chief Strategy Officer Jason Kwon reportedly conveyed this sentiment to the staff in a memo on Saturday night, as per sources familiar with the matter. The initial decision to terminate Altman, citing a lack of candor with the board, could undergo a surprising turnaround. The escalating employee departures and internal unrest, marked by discussions of a "hostile takeover," have exerted significant pressure on the board, prompting a reevaluation of their stance.
According to Gartner, a market research and information technology consulting firm, the global cloud market is projected to reach $678.8 billion in 2024, marking a substantial 20.4% increase from the $563.6 billion recorded in the current year. The segment anticipated to experience the most significant growth is cloud infrastructure, the foundational platform supporting AI, software, and applications, with an expected growth rate of 26.6% next year. Major players in the cloud market, including Amazon, Microsoft, and Google, are turning to generative AI to revitalize their cloud businesses. This strategic move comes at a time when business leaders are scaling back on software spending and seeking ways to manage cloud costs effectively. However, despite the enthusiasm surrounding artificial intelligence, tech vendors have faced challenges in translating the AI hype into profits. Many Chief Information Officers (CIOs) and consumers are still in the process of determining how they will leverage AI and what they are willing to pay for such services, contributing to the industry's struggle to convert AI investments into substantial financial gains.
Amazon has announced the elimination of "several hundred" positions associated with its Alexa smart assistant, according to a memo sent to employees. The e-commerce giant stated that it is realigning its focus by shifting efforts towards generative AI to better align with its business priorities. The job cuts are occurring in the U.S., Canada, and India, and the company is also exploring reductions in headcount in other regions. Amazon is providing affected employees with severance packages that include a separation payment, transitional health insurance benefits, and paid time for job searches. Daniel Rausch, the Vice President of Alexa and Fire TV, emphasized that despite the challenging decision, Amazon remains optimistic about the future of Alexa, emphasizing its continued importance to the company's overall business.
Alibaba announced on Thursday that it has abandoned its initiative to spin off and list its cloud-computing division, citing the repercussions of export controls implemented late last month. The company expressed concerns that these restrictions could significantly and negatively impact the cloud business's capacity to provide products and services, as well as fulfill existing contracts. In light of these challenges, Alibaba has decided to concentrate on fostering the growth of the cloud division. The original intention was to separate the cloud business and aim for a listing by May 2024, as part of a broader historic restructuring plan designed to divide the company into six main business groups. However, the recent developments in export controls have prompted a shift in strategy for Alibaba.
Google's ambitious initiative to catch up with OpenAI, the creator of ChatGPT, appears to be facing unexpected challenges. Earlier in the year, Google representatives informed some cloud customers and business partners that they would gain access to the company's new conversational AI, a substantial language model known as Gemini, by November. However, recent reports indicate that Google has notified these parties to expect a delay until the first quarter of the following year, as reported by The Information. This setback comes amid slowing cloud sales growth for Google compared to its rival Microsoft, which has been leveraging OpenAI's technology. The delay suggests Google may prioritize strengthening consumer offerings before giving external developers access. Sundar Pichai hinted at laying the foundation for next-gen models during an investor call on October 24.
Meta Platforms has undergone organizational changes, splitting a team responsible for understanding and preventing harms associated with its developing artificial intelligence (AI) technology. The company is redirecting resources toward its generative AI work, as revealed in an internal post reviewed by The Information. The reshuffling of AI teams at Meta involves most employees from the Responsible AI team transitioning to Meta's generative AI team, established in February to focus on generative AI products. Additionally, some members will join Meta's AI infrastructure unit, which is dedicated to developing systems and tools for building and operating AI products.
Amazon Web Services (AWS) has introduced PartyRock, a consumer service aimed at individuals without software development experience, enabling them to create generative artificial intelligence applications. The focus of PartyRock, seemingly designed for a Gen Z audience, is to facilitate learning about interacting with generative AI. However, it is explicitly stated to be for "entertainment purposes only" and is not intended for enterprise customers. In the enterprise sales realm, AWS has been striving to catch up with competitors like Microsoft and OpenAI. Users of PartyRock can build applications for various tasks, such as generating a recipe for a three-course meal to complement a bottle of wine or creating a music playlist based on individual preferences. The product is based on Amazon Bedrock, a cloud service that allows developers to work with large language models through an application programming interface (API). Initially, PartyRock users can build apps using models from Anthropic, AI21 Labs, Cohere, and Stability AI. AWS plans to incorporate more language model options over time as they become available in Bedrock. Notably, the PartyRock website does not mention Titan, AWS's own large language models.
Tencent Holdings Ltd. has addressed concerns about potential constraints on its AI development capabilities due to US export controls. The company, based in Shenzhen, stated that it has amassed a significant inventory of Nvidia Corp.'s H800 artificial intelligence accelerators. This stockpile is substantial enough to support the development of its proprietary Hunyuan AI model for at least a couple more generations, according to Tencent President Martin Lau. Lau emphasized that Tencent currently holds one of the largest inventories of AI chips in China, having been among the first to place orders for the H800, ensuring a robust supply for ongoing development efforts.
Microsoft has introduced its first in-house artificial intelligence chip called Maia 100, along with a cloud-computing processor named Cobalt. This move is aimed at increasing control over its technology and enhancing its offerings in the competitive AI computing market. Unveiled at the company's annual Ignite conference in Seattle, Maia 100 is designed to enable Microsoft Azure cloud customers to develop and execute AI programs for content generation. The chip is currently being tested with Microsoft's Bing and Office AI products. Microsoft's primary AI partner, OpenAI (creator of ChatGPT), is also involved in testing the Maia 100 processor. Both Maia and the server chip, Cobalt, are set to debut in select Microsoft data centers early next year. Additionally, Microsoft announced new software allowing clients to design their own AI assistants.
Apple Inc. is facing further setbacks in its efforts to produce an in-house modem chip for the iPhone, experiencing challenges in replacing the intricate Qualcomm Inc. component. Originally planning to have its own chip ready by next year, Apple is now expected to miss a target to ship the component by the spring of 2025, according to sources familiar with the matter. This delay could push the release to at least the end of 2025 or early 2026, coinciding with the final year of Apple's extended contract with Qualcomm. The difficulties highlight the complexity of designing a modem that seamlessly connects to global cellular networks, meets performance standards set by industry leaders like Qualcomm, and functions reliably across diverse environments and conditions.
Starting next year, Amazon customers will have the option to purchase Hyundai Motor Co. vehicles directly through the e-commerce platform. Currently, while shoppers can choose cars from various automakers on Amazon, they need to connect with a dealer to complete the transaction. The new arrangement with Hyundai will allow customers to buy a Hyundai vehicle directly from the dealer on Amazon, marking a significant shift in the online car-buying experience. This partnership, the first of its kind between Amazon and an automaker, could have broader implications, potentially prompting other car manufacturers to explore similar collaborations. The move comes as the automotive industry increasingly embraces digital channels, particularly amid a shift to online vehicle purchases and the growing focus on electric vehicles. Despite these shifts, strict franchise laws in the U.S. have traditionally maintained dealers as intermediaries in new-car sales transactions.
Semiconductor startup Sapeon Inc., backed by South Korean conglomerate SK Group, has introduced its latest artificial intelligence chip, the Sapeon X330. The new chip boasts four times the computational performance and over twice the power efficiency compared to its predecessor, the X220. Sapeon, based in Santa Clara, California, has received backing from major shareholders including SK Telecom Co., SK Hynix Inc., and investment firm SK Square Co. The company plans to conduct testing for significant customers before commencing mass production in the first half of 2024. This move is part of Sapeon's efforts to strengthen its position in the competitive AI chip market.
It was recently reported that Amazon is working on developing its own operating system, codenamed Vega, to replace Android on its Fire TV devices and potentially extend to other products, including smart displays. Originally expected to debut on Fire TV streaming hardware next year, it has now been discovered that this new OS is already in use on the latest Echo Show 5 devices. The software on the Echo Show 5 is labeled as "OS 1.1," a significant departure from the expected Fire OS designation. Despite this change, the Vega OS retains a visual and user experience similarity to the Fire OS it is replacing.
Google has filed a lawsuit against unnamed individuals in India and Vietnam, alleging that scammers are exploiting the surge in consumer interest in artificial intelligence tools to pilfer social media account passwords of U.S. small businesses. The lawsuit, filed on Monday, claims that hackers have been deceiving small-business owners by enticing them to click on Facebook ads that promise to download Google's Bard artificial intelligence chatbot. However, the ads are deceptive, as Bard is a free web-based platform and not available for download. The scammers use these false ads to deliver malware that steals social media credentials. The organized group of hackers operates through accounts and pages with misleading names like Google AI, AIGoogle.Plus, AIGoogle Bard FB, and AIGoogleBard, according to Google.
In the ongoing antitrust trial, it was revealed that Google pays Apple Inc. a substantial 36% of the revenue generated from search advertising through the Safari browser. The disclosure came from Kevin Murphy, a University of Chicago professor serving as the main economics expert for Google, during his testimony. The revelation, which was intended to remain confidential, visibly surprised Google's main litigator, John Schmidtlein. Both Google and Apple had objected to making details of their agreement public, with Google arguing that revealing additional information about the deal could unreasonably undermine its competitive standing.
In a significant shift, Apple Inc. has announced plans to adopt the RCS (Rich Communication Services) standard next year, allowing smoother text messaging between iPhones and Android devices. The move comes after Apple resisted adopting RCS for over a year, despite pressure from Google and other entities advocating for the technology. RCS is an enhanced messaging standard backed by the GSM Association, providing improvements over standard SMS and MMS texting. With this adoption, users will enjoy additional features such as the ability to text between platforms over Wi-Fi, send larger video and photo files, engage in more efficient group chats, and check if messages have been received and read.
Thanks for reading Uncover Tech Insights! Subscribe for free to receive new posts and support my work.
We would really appreciate it if you could share it with someone you know who would be interested in tech investing news. Organic, and especially word of mouth, is the main way our newsletter will grow. We will always put quality before quantity. Thank you.
Source: the expert interviews are found on Alphasense, a platform specialized for expert interviews and other investment data.
Nothing contained in this website and newsletter should be understood as investment or financial advice. All investment strategies and investments involve the risk of loss. Past performance does not guarantee future results. Everything written and expressed in this newsletter is only the writer's opinion and should not be considered investment advice. Before investing in anything, know your risk profile and if needed, consult a professional. Nothing on this site should ever be considered advice, research, or an invitation to buy or sell any securities.