Connect with us

Published

on

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. 

Justin Sullivan | Getty Images News | Getty Images

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.”

— CNBC’s Jordan Novet contributed to this report.

Don’t miss these stories from CNBC PRO:

Continue Reading

Technology

These underperforming groups may deliver AI-electric appeal. Here’s why.

Published

on

By

These underperforming groups may deliver AI-electric appeal. Here's why.

Reshoring and infrastructure products could be the next ETF play after AI, say ETF experts

Industrial and infrastructure stocks may soon share the spotlight with the artificial intelligence trade.

According to ETF Action’s Mike Atkins, there’s a bullish setup taking shape due to both policy and consumer trends. His prediction comes during a volatile month for Big Tech and AI stocks.

“You’re seeing kind of the old-school infrastructure, industrial products that have not done as well over the years,” the firm’s founding partner told CNBC’s “ETF Edge” this week. “But there’s a big drive… kind of away from globalization into this reshoring concept, and I think that has legs.”

Global X CEO Ryan O’Connor is also optimistic because the groups support the AI boom. His firm runs the Global X U.S. Infrastructure Development ETF (PAVE), which tracks companies involved in construction and industrial projects.

“Infrastructure is something that’s near and dear to our heart based off of PAVE, which is our largest ETF in the market,” said O’Connor in the same interview. “We think some of these reshoring efforts that you can get through some of these infrastructure places are an interesting one.”

The Global X’s infrastructure exchange-traded fund is up 16% so far this year, while the VanEck Semiconductor ETF (SMH), which includes AI bellwethers Nvidia, Taiwan Semiconductor and Broadcom, is up 42%, as of Friday’s close.

Both ETFs are lower so far this month — but Global X’s infrastructure ETF is performing better. Its top holdings, according to the firm’s website, are Howmet Aerospace, Quanta Services and Parker Hannifin.

Supporting the AI boom

He also sees electrification as a positive driver.

“All of the things that are going to be required for us to continue to support this AI boom, the electrification of the U.S. economy, is certainly one of them,” he said, noting the firm’s U.S. Electrification ETF (ZAP) gives investors exposure to them. The ETF is up almost 24% so far this year.

The Global X U.S. Electrification ETF is also performing a few percentage points better than the VanEck Semiconductor ETF for the month.

Disclaimer

Continue Reading

Technology

How tariffs and AI are giving secondhand platforms like ThredUp a boost

Published

on

By

How tariffs and AI are giving secondhand platforms like ThredUp a boost

At ThredUp‘s 600,000-square-foot warehouse in Suwanee, Georgia, roughly 40,000 pieces of used clothing are processed each day. The company’s logistics network — four facilities across the U.S. — now rivals that of some fast-fashion giants.

“This is the largest garment-on-hanger system in the world,” said Justin Pina, ThredUp’s senior director of operations. “We can hold more than 3.5 million items here.”

Secondhand shopping is booming. The global secondhand apparel market is expected to reach $367 billion by 2029, growing almost three times faster than the overall apparel market, according to GlobalData.

President Donald Trump’s tariffs were billed as a way to bring manufacturing back home. But the measures hit one of America’s most import-dependent industries: fashion.

About 97 percent of clothing sold in the U.S. is imported, mostly from China, Vietnam, Bangladesh and India, according to the American Apparel and Footwear Association.

For years, Gen Z shoppers have been driving the rise of secondhand fashion, but now more Americans are catching on.

“When tariffs raise those costs, resale platforms suddenly look like the smart buy. This isn’t just a fad,” said Jasmine Enberg, co-CEO of Scalable. “Tariffs are accelerating trends that were already reshaping the way Americans shop.”

For James Reinhart, ThredUp’s CEO, the company is already seeing it play out.

“The business is free-cash-flow positive and growing double digits,” said Reinhart. “We feel really good about the economics, gross margins near 80% and operations built entirely within the U.S.”

ThredUp reported that revenue grew 34% year over year in the third quarter. The company also said it acquired more new customers in the quarter than at any other time in its history, with new buyer growth up 54% from the same period last year.

“If tariffs add 20% to 30% to retail prices, that’s a huge advantage for resale,” said Dylan Carden, research analyst at William Blair & Company. “Pre-owned items aren’t subject to those duties, so demand naturally shifts.”

Inside the ThredUp warehouse, where CNBC got a behind-the-scenes look. automation hums alongside human workers. AI systems photograph, categorize, and price thousands of garments per hour. For Reinhart, the technology is key to scaling resale like retail.

“AI has really accelerated adoption,” said Reinhart. “It’s helping us improve discovery, styling, and personalization for buyers.”

That tech wave extends beyond ThredUp. Fashion-tech startups Phia, co-founded by Phoebe Gates and Sophia Kianni, is using AI to scan thousands of listings across retail and resale in seconds.

“The fact that we’ve driven millions in transaction volume shows how big this need is,” Gates said. “People want smarter, cheaper ways to shop.”

ThredUp is betting that domestic infrastructure, automation, and AI will keep it ahead of the curve, and that tariffs meant to revive U.S. manufacturing could end up powering a new kind of American fashion economy.

“The future of fashion will be more sustainable than it is today,” said Reinhart. “And secondhand will be at the center of it.”

Watch the video to learn more.

Continue Reading

Technology

AI anxiety on the rise: Startup founders react to bubble fears

Published

on

By

AI anxiety on the rise: Startup founders react to bubble fears

Markets were on edge this week as a steady stream of negative headlines around the artificial intelligence trade stoked fears of a bubble.

Famed short-seller Michael Burry cast doubt on the sustainability of AI earnings. Concerns around the levels of debt funding AI infrastructure buildouts grew louder. And once high-flyers like CoreWeave tanked on disappointing guidance.

CNBC’s Deirdre Bosa asked those at the epicenter of the boom for their take, sitting down with the founders of two of the buzziest AI startups.

Amjad Masad, founder and CEO of AI coding startup Replit, admits there’s been a cooldown.

“Early on in the year, there was the vibe coding hype market, where everyone’s heard about vibe coding. Everyone wanted to go try it. The tools were not as good as they are today. So I think that burnt a lot of people,” Masad said. “So there’s a bit of a vibe coding, I would say, hype slow down, and a lot of companies that were making money are not making as much money.”

Masad added that a lot companies were publishing their annualized recurring revenue figures every week, and “now they’re not.”

Navrina Singh, founder and CEO of startup Credo AI, which helps enterprises with AI oversight and risk management, is seeing more excitement than fear.

“I don’t think we are in a bubble,” she said. “I really believe this is the new reality of the world that we are living in. As we know, AI is going to be and already is our biggest growth driver for businesses. So it just makes sense that there has to be more investment, not only on the capability side, governance side, but energy and infrastructure side as well.”

Watch this video to learn more. 

Continue Reading

Trending