Alphabet offers reasons to believe it is closing the gap in the AI race

Original Article:

Source: Link

Alphabet 's cloud-computing summit left us feeling much better about the technology giant's artificial intelligence efforts. There was a lot riding on the event , considering shares hovered near all-time highs heading into it and our general frustration around its AI execution, but the Club holding cleared our bar. Alphabet's announcements Tuesday at Google Cloud Next including a new server chip and several generative AI service offerings — alongside favorable commentary from key Google Cloud partners — made it clear that counting the company out of the competitive AI race is foolish. We'd never gone that far, but the search engine heavyweight certainly had tested our patience since the launch of Microsoft-based ChatGPT in late 2022 took the investment world by storm and established generative AI as a household term. Long thought to be an AI leader, Alphabet was suddenly playing catchup to Microsoft. Its shortcomings with the technology had continued into this year with its image-generation controversy. We trimmed our position twice in 2024, most recently on April 1 . Our focus is now on where the company and its stock go from here. In his presentation at the event, Google Cloud CEO Thomas Kurian showed us that Alphabet is not content with being anything other than a leader and innovator in the rapidly growing area of AI. Our hope is he's right. Shares of Alphabet fell 0.3% on Wednesday, breaking a three-day win streak, though they still held up better much better than the broader market. The stock is up nearly 12% year to date. AI Agents The launch of "AI Agents" demonstrated the real-world applications of Alphabet's generative AI advancements — a key box that investors want to see checked in the search for money-making opportunities . It's fun to play around with generative AI applications like OpenAI's ChatGPT and Alphabet's Gemini, but, at this point in the cycle, investors want clarity on how companies plan to drive mass adoption and monetize the titanic amounts of money being invested in AI infrastructure and research budgets to develop ever-more-capable large language models , or LLMs. Alphabet's agents appear to be similar to Microsoft's Copilot virtual assistants, which rolled out to customers last year , and Nvidia 's "NIMs," which the leading AI chipmaker unveiled at its developer conference in March . These AI tools are designed for specific use cases and can help customers or employees in either administrative or creative tasks, as well as provide a hand on the back end with processing data, writing software code and improving security. These applications are clearly the next phase of the generative AI revolution beyond the development of LLMs and deployment of general-purpose chatbots. Companies that can best leverage their LLMs to provide best-in-class agents — or Copilots or NIMs — will thrive. So, too, will their customers that harness the power of the AI agents to better serve their own clients, increase employee productivity, and drive corporate efficiencies. GOOGL 1Y mountain Alphabet's stock performance over the past 12 months. Updates to Vertex AI Alphabet updated Vertex AI — its platform for building and deploying AI models — to include an agent-building feature. That essentially allows businesses to create customized AI agents on top of Gemini models. Another applause-worthy improvement to VertexAI is the ability to "ground" AI models with Google Search. Essentially, this connects AI models to verifiable sources of information through Google Search and, as a result, "significantly" improves the accuracy of responses, according to Alphabet. One issue with generative AI models is hallucinations , in which the model provides responses to user inquiries that contain false information, even though they are presented as if they are fact. The effect of both announcements is fortifying Google Cloud's standing among the cloud providers, which are competing to attract companies and developers who want tools to advance their AI strategies. Google Cloud is the third-largest cloud provider behind Amazon Web Services and Microsoft's Azure. Alphabet also said Gemini 1.5 Pro is now in public preview, making the model available to developers via Vertex AI. A key feature of the model is its massive context window compared with prior generations and offerings from rivals. This gives Gemini the ability to summarize larger text documents and scrutinize bigger blocks of code, among other advancements. The model also has improved audio capabilities. In addition to being able to summarize an audio file, a user could ask the model to find where the soundbite "play ball" is said and in response receive a time stamp for that occurrence, Kurian explained. The updates to Gemini highlight the company's ability to innovate within AI, despite the perception issues it's faced over the past year and a half. Advancements in hardware Alphabet's custom hardware lineup is growing with the introduction of Google Axion, its first Arm-based central processing unit designed for the data center. Google Cloud instances using the chip will be available to customers later this year. For many years, the company has developed in-house AI chips, known as Tensor Processing Units, and the most powerful version of the TPU is now generally available. But the Axion chip takes its custom silicon journey into data-center terrain long dominated by Intel and Advanced Micro Devices. Amazon and Microsoft also have developed custom CPUs based on Arm architecture. Alphabet said the Axion chip delivers performance improvements over the "fastest general-purpose Arm-based instances available in the cloud today," as well as performance and energy-efficiency gains over CPUs that use the x86 architecture favored by Intel and AMD. Elsewhere, the company also confirmed that Nvidia's latest generation of chips, known as Blackwell, will be available to Google Cloud customers starting in 2025. Bottom Line The announcements at Google Cloud Next suggest that Alphabet's rocky start in the generative AI race has not impaired its ability to innovate and make progress on the technology. The AI revolution remains in its early stages, but Kurian and Co. at Google Cloud leave us feeling very good about both its ability to compete on AI offerings and remain a top-three cloud provider in the next generation of computing. (Jim Cramer's Charitable Trust is long MSFT See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED. 'Welcome to the Gemini era' slogan on Google DeepMind website is screened on a mobile phone for illustration photo. Gliwice,Poland on December 27, 2023. Beata Zawrzel | Nurphoto | Getty Images