Google CEO Sundar Pichai speaks in conversation with Emily Chang during the APEC CEO Summit at Moscone West on November 16, 2023 in San Francisco, California. The APEC summit is being held in San Francisco and runs through November 17.
Google is launching what it considers its largest and most capable artificial intelligence model Wednesday as pressure mounts on the company to answer how it’ll monetize AI.
The large language model Gemini will include a suite of three different sizes: Gemini Ultra, its largest, most capable category; Gemini Pro, which scales across a wide range of tasks; and Gemini Nano, which it will use for specific tasks and mobile devices.
For now, the company is planning to license Gemini to customers through Google Cloud for them to use in their own applications. Starting Dec. 13, developers and enterprise customers can access Gemini Pro via the Gemini API in Google AI Studio or Google Cloud Vertex AI. Android developers will also be able to build with Gemini Nano. Gemini will also be used to power Google products like its Bard chatbot and Search Generative Experience, which tries to answer search queries with conversational-style text (SGE is not widely available yet).
Gemini Ultra is the first model to outperform human experts on MMLU (massive multitask language understanding), which uses a combination of 57 subjects such as math, physics, history, law, medicine and ethics for testing both world knowledge and problem-solving abilities, the company said in a blog post Wednesday. It can supposedly understand nuance and reasoning in complex subjects.
Sundar Pichai, chief executive officer of Alphabet Inc., during the Google I/O Developers Conference in Mountain View, California, US, on Wednesday, May 10, 2023.
David Paul Morris | Bloomberg | Getty Images
“Gemini is the result of large-scale collaborative efforts by teams across Google, including our colleagues at Google Research,” wrote CEO Sundar Pichai in a blog post Wednesday. “It was built from the ground up to be multimodal, which means it can generalize and seamlessly understand, operate across and combine different types of information including text, code, audio, image and video.”
Starting today, Google’s chatbot Bard will use Gemini Pro to help with advanced reasoning, planning, understanding and other capabilities. Early next year, it will launch “Bard Advanced,” which will use Gemini Ultra, executives said on a call with reporters Tuesday. It represents the biggest update to Bard, its ChatGPT-like chatbot.
The update comes eight months after the search giant first launched Bard and one year after OpenAI launched ChatGPT on GPT-3.5. In March of this year, the Sam Altman-led startup launched GPT-4. Executives said Tuesday that Gemini Pro outperformed GPT-3.5 but dodged questions about how it stacked up against GPT-4.
When asked if Google has plans to charge for access to “Bard Advanced,” Google’s general manager for Bard, Sissie Hsiao, said it is focused on creating a good experience and doesn’t have any monetization details yet.
When asked on a press briefing if Gemini has any novel capabilities compared with current generation LLMs, Eli Collins, vice president of product at Google DeepMind, answered, “I suspect it does” but that it’s still working to understand Gemini Ultra’s novel capabilities.
Google reportedly postponed the launch of Gemini because it wasn’t ready, bringing back memories of the company’s rocky rollout of its AI tools at the beginning of the year.
Multiple reporters asked about the delay, to which Collins answered that testing the more advanced models take longer. Collins said Gemini is the most highly tested AI model that the company’s built and that it has “the most comprehensive safety evaluations” of any Google model.
Collins said that despite being its largest model, Gemini Ultra is significantly cheaper to serve. “It’s not just more capable, it’s more efficient,” he said. “We still require significant compute to train Gemini but we’re getting much more efficient in terms of our ability to train these models.”
Collins said the company will release a technical white paper with more details of the model on Wednesday but said it won’t be releasing the perimeter count. Earlier this year, CNBC found Google’s PaLM 2 large language model, its latest AI model at the time, used nearly five times the amount of text data for training as its predecessor LLM.
Also on Wednesday, Google introduced its next-generation tensor processing unit for training AI models. The TPU v5p chip, which Salesforce and startup Lightricks have begun using, offers better performance for the price than the TPU v4 announced in 2021, Google said. But the company didn’t provide information on performance compared with market leader Nvidia.
The chip announcement comes weeks after cloud rivals Amazon and Microsoft showed off custom silicon targeting AI.
During Google’s third-quarter earnings conference call in October, investors asked executives more questions about how it’s going to turn AI into actual profit.
In August, Google launched an “early experiment” called Search Generative Experience, or SGE, which lets users see what a generative AI experience would look like when using the search engine — search is still a major profit center for the company. The result is more conversational, reflecting the age of chatbots. However, it is still considered an experiment and has yet to launch to the general public.
Investors have been asking for a timeline for SGE since May, when the company first announced the experiment at its annual developer conference Google I/O. The Gemini announcement Wednesday hardly mentioned SGE and executives were vague about its plans to launch to the general public, saying that Gemini would be incorporated into it “in the next year.”
“This new era of models represents one of the biggest science and engineering efforts we’ve undertaken as a company,” Pichai said in Wednesday’s blog post. “I’m genuinely excited for what’s ahead, and for the opportunities Gemini will unlock for people everywhere.”
Artificial intelligence robot looking at futuristic digital data display.
Yuichiro Chino | Moment | Getty Images
Businesses are turning to artificial intelligence tools to help them navigate real-world turbulence in global trade.
Several tech firms told CNBC say they’re deploying the nascent technology to visualize businesses’ global supply chains — from the materials that are used to form products, to where those goods are being shipped from — and understand how they’re affected by U.S. President Donald Trump’s reciprocal tariffs.
Last week, Salesforce said it had developed a new import specialist AI agent that can “instantly process changes for all 20,000 product categories in the U.S. customs system and then take action on them” as needed, to help navigate changes to tariff systems.
Engineers at the U.S. software giant used the Harmonized Tariff Schedule, a 4,400-page document of tariffs on goods imported to the U.S., to inform answers generated by the agent.
“The sheer pace and complexity of global tariff changes make it nearly impossible for most businesses to keep up manually,” Eric Loeb, executive vice president of government affairs at Salesforce, told CNBC. “In the past, companies might have relied on small teams of in-house experts to keep pace.”
Firms say that AI systems are enabling them to take decisions on adjustments to their global supply chains much faster.
Andrew Bell, chief product officer of supply chain management software firm Kinaxis, said that manufacturers and distributors looking to inform their response to tariffs are using his firm’s machine learning technology to assess their products and the materials that go into them, as well as external signals like news articles and macroeconomic data.
“With that information, we can start doing some of those simulations of, here is a particular part that is in your build material that has a significant tariff. If you switched to using this other part instead, what would the impact be overall?” Bell told CNBC.
‘AI’s moment to shine’
Trump’s tariffs list — which covers dozens of countries — has forced companies to rethink their supply chains and pricing, with the likes of Walmart and Nikealready raising prices on some products. The U.S. imported about $3.3 trillion of goods in 2024, according to census data.
Uncertainty from the U.S. tariff measures “actually probably presents AI’s moment to shine,” Zack Kass, a futurist and former head of OpenAI’s go-to-market strategy, told CNBC’s Silvia Amaro at the Ambrosetti Forum in Italy last month.
Read more CNBC tech news
“If you wonder how hard things could get without AI vis-a-vis automation, and what would happen in a world where you can’t just employ a bunch of people overnight, AI presents this alternative proposal,” he added.
Nagendra Bandaru, managing partner and global head of technology services at Indian IT giant Wipro, said clients are using the company’s agentic AI solutions “to pivot supplier strategies, adjust trade lanes, and manage duty exposure dynamically as policy landscapes evolve.”
Wipro says it uses a range of AI systems — both proprietary and supplied by third parties — from large language models to traditional machine learning and computer vision techniques to inspect physical assets in cross-border transit.
‘Not a silver bullet’
While it preferred to keep company names confidential, Wipro said that firms using its AI products to navigate Trump’s tariffs range from a Fortune 500 electronics manufacturer with factories in Asia to an automotive parts supplier exporting to Europe and North America.
“AI is a powerful enabler — but not a silver bullet,” Bandaru told CNBC. “It doesn’t replace trade policy strategy, it enhances it by transforming global trade from a reactive challenge into a proactive, data-driven advantage.”
AI was already a key investment priority for global firms prior to Trump’s sweeping tariff announcements on April. Nearly three-quarters of business leaders ranked AI and generative AI in their top three technologies for investment in 2025, according to a report by Capgemini published in January.
“There are a number of ways AI can assist companies dealing with the tariffs and resulting uncertainty. But any AI solution’s success will be predicated on the quality of the data it has access to,” Ajay Agarwal, partner at Bain Capital Ventures, told CNBC.
The venture capitalist said that one of his portfolio companies, FourKites, uses supply chain network data with AI to help firms understand the logistics impacts of adjusting suppliers due to tariffs.
“They are working with a number of Fortune 500 companies to leverage their agents for freight and ocean to provide this level of visibility and intelligence,” Agarwal said.
“Switching suppliers may reduce tariffs costs, but might increase lead times and transportation costs,” he added. “In addition, the volatility of the tariffs [has] severely impacted the rates and capacity available in both the ocean and the domestic freight networks.”
A Zoox autonomous robotaxi in San Francisco, California, US, on Wednesday, Dec. 4, 2024.
David Paul Morris | Bloomberg | Getty Images
Amazon‘s Zoox robotaxi unit issued a voluntary recall of its software for the second time in a month following a recent crash in San Francisco.
On May 8, an unoccupied Zoox robotaxi was turning at low speed when it was struck by an electric scooter rider after braking to yield at an intersection. The person on the scooter declined medical attention after sustaining minor injuries as a result of the collision, Zoox said.
“The Zoox vehicle was stopped at the time of contact,” the company said in a blog post. “The e-scooterist fell to the ground directly next to the vehicle. The robotaxi then began to move and stopped after completing the turn, but did not make further contact with the e-scooterist.”
Zoox said it submitted a voluntary software recall report to the National Highway Traffic Safety Administration on Thursday.
A Zoox spokesperson said the notice should be published on the NHTSA website early next week. The recall affected 270 vehicles, the spokesperson said.
The NHTSA said in a statement it had received the recall notice and that the agency “advises road users to be cautious in the vicinity of vehicles because drivers may incorrectly predict the travel path of a cyclist or scooter rider or come to an unexpected stop.”
If an autonomous vehicle continues to move after contact with any nearby vulnerable road user, it risks causing harm or further harm. In the AV industry, General Motors-backed Cruise exited the robotaxi business after a collision in which one of its vehicles injured a pedestrian who had been struck by a human-driven car and was then rolled over by the Cruise AV.
Zoox’s May incident comes roughly two weeks after the company announced a separate voluntary software recall following a recent Las Vegas crash. In that incident, an unoccupied Zoox robotaxi collided with a passenger vehicle, resulting in minor damage to both vehicles.
The company issued a software recall for 270 of its robotaxis in order to address a defect with its automated driving system that could cause it to inaccurately predict the movement of another car, increasing the “risk of a crash.”
Amazon acquired Zoox in 2020 for more than $1 billion, announcing at the time that the deal would help bring the self-driving technology company’s “vision for autonomous ride-hailing to reality.”
While Zoox is in a testing and development stage with its AVs on public roads in the U.S., Alphabet’s Waymo is already operating commercial, driverless ride-hailing services in Phoenix, San Francisco, Los Angeles and Austin, Texas, and is ramping up in Atlanta.
Teslais promising it will launch its long-delayed robotaxis in Austin next month, and, if all goes well, plans to expand after that to San Francisco, Los Angeles and San Antonio, Texas.
Shares of Intuit popped about 9% on Friday, a day after the company reported quarterly results that beat analysts’ estimates and issued rosy guidance for the full year.
Intuit, which is best known for its TurboTax and QuickBooks software, said revenue in the fiscal third quarter increased 15% to $7.8 billion. Net income rose 18% to $2.82 billion, or $10.02 per share, from $2.39 billion, or $8.42 per share, a year earlier.
“This is the fastest organic growth that we have had in over a decade,” Intuit CEO Sasan Goodarzi told CNBC’s “Closing Bell: Overtime” on Thursday. “It’s really incredible growth across the platform.”
For its full fiscal year, Intuit said it expects to report revenue of $18.72 billion to $18.76 billion, up from the range of $18.16 billion to $18.35 billion it shared last quarter. Analysts were expecting $18.35 billion, according to LSEG.
“We’re redefining what’s possible with [artificial intelligence] by becoming a one-stop shop of AI-agents and AI-enabled human experts to fuel the success of consumers and small and mid-market businesses,” Goodarzi said in a release Thursday.
Read more CNBC tech news
Goldman Sachs analysts reiterated their buy rating on the stock and raised their price target to $860 from $750 on Thursday. The analysts said Intuit’s execution across its core growth pillars is “reinforcing confidence” in its growth profile over the long term.
The company’s AI roadmap, which includes the introduction of AI agents, will add additional upside, the analysts added.
“In our view, Intuit stands out as a rare asset straddling both consumer and business ecosystems, all while supplemented by AI-prioritization,” the Goldman Sachs analysts wrote in a note.
Analysts at Deutsche Bank also reiterated their buy rating on the stock and raised their price target to $815 from $750.
They said the company’s results were “reassuring” after a rocky two years and that they feel more confident about its ability to grow the consumer business.
“Longer term, we continue to believe Intuit presents a unique investment opportunity and we see its platform approach powering accelerated innovation with leverage, thus enabling sustained mid-teens or better EPS growth,” the analysts wrote in a Friday note.