Zoho’s annual user and partner conference ‘Zoholics’ offers SMBs an ‘on ramp’ to applying AI in their business, Researchers claim 95% reduction in neural network energy consumption, DOJ considers breaking up Google after monopoly ruling, AI Pioneer Geoffrey Hinton wins Nobel Prize, takes a jab at OpenAI’s CEO
Welcome to Hashtag Trending. I’m your host, Jim Love. Let’s get into it.
Zoho’s annual user and partner conference ‘Zoholics’ offers SMBs an ‘on ramp’ to applying AI in their business.
During a briefing for press and analysts, Zoho’s Canadian Managing Director, Chandrashekar LSP, highlighted the company’s growth in Canada. While its primary growth in the region has been among small and medium-sized businesses (SMBs), Zoho is also making significant inroads into enterprise accounts. The company now boasts over 35 employees in Canada, most of whom are bilingual, along with a growing number of dedicated channel partners. With its own Canadian data center, Zoho appears poised for significant expansion in the coming years.
Zoho, essentially an ERP system, offers over 45 different modules that can be purchased individually or integrated to allow seamless data sharing across the entire suite. This integration gives Zoho an advantage in analytics, even for smaller businesses, and provides a unique platform to bring artificial intelligence to the SMB market, reducing adoption barriers.
One challenge for SMBs is knowing where to start with AI. To address this, Zoho’s AI Evangelist, Ramprakash Ramamoorthy, unveiled a suite of ‘narrow AI’ agents—AI-driven applications targeting specific use cases in areas such as sales, marketing, customer service, finance, HR, IT operations, legal compliance, and facilities management. By providing a portfolio of AI applications, Zoho aims to make it easier for smaller businesses to achieve early, realistic results in their AI initiatives.
Zoho supports a choice of large language models (LLMs) like ChatGPT, ClaudeAI, and Llama, while also developing its own LLM. Thanks to the integrated data structure of Zoho’s suite, even modest-sized businesses can leverage AI across their company data, which remains securely hosted within Zoho’s ecosystem.
Zoho has consistently differentiated itself with a highly competitive price point. One customer noted they were able to access the entire suite of products for about the same cost as a single CRM application from other vendors. This value extends to Zoho’s AI offerings as well.
In addition to price, Zoho is also emphasizing privacy as a differentiator. Chandrashekar LSP emphasized that Zoho does not use customer information to train its AI models. By combining cost advantages, privacy assurances, and a clear path for SMBs to start and grow their AI usage, Zoho aims to capitalize on the rising interest in generative AI among small businesses.
Sources: Zoho
Researchers claim 95% reduction in neural network energy consumption
Many of us feel conflicted about our enthusiasm about AI and our desire for a more sustainable footprint for IT.
A team of researchers has unveiled a new technique called L-Mul, which they say could reduce energy consumption in large language models (LLMs) by up to 95%. This breakthrough addresses one of the key challenges of artificial intelligence today: the energy-intensive nature of floating-point multiplications used in neural networks.
The problem of high energy usage has become increasingly significant as AI models continue to grow in scale and complexity. Current LLMs often rely on billions of floating-point operations, which require enormous amounts of electricity. L-Mul, short for “Light Multiplication,” aims to reduce the need for traditional floating-point multiplications by employing alternative arithmetic methods that are far more efficient.
According to researchers, L-Mul not only dramatically cuts down on power usage but also maintains the accuracy and speed needed for LLMs to perform effectively. This advancement could have wide-ranging implications for the development of AI, especially in industries where computational costs and environmental concerns are critical.
The introduction of L-Mul might not just lower the energy footprint of AI but could also make sophisticated AI more accessible to companies with limited computational resources, democratizing access to advanced models and opening doors for broader applications in sectors like healthcare, education, and climate modeling.
#Sources Include:# Analytics India Magazine
And https://arxiv.org/abs/2410.00907
DOJ considers breaking up Google after monopoly ruling
The U.S. Department of Justice (DOJ) is considering a breakup of Google as a remedy for the company’s illegal monopoly in online search. This comes after a landmark court ruling in August found that Google had unlawfully maintained its dominance through a series of self-reinforcing tactics, making it difficult for competitors to establish a foothold in the market. The DOJ suggested possible remedies in a court filing on Tuesday, stating that “remedying Google’s unlawful conduct takes time” and may involve significant measures.
Breaking up Google, the world’s largest search engine, would be the most consequential antitrust action in the U.S. in decades. Since the ruling, federal Judge Amit Mehta has been weighing options, but a decision to split up the company could be difficult to implement. The case centers on Google’s contracts with companies like Apple and Mozilla, which made Google the default search engine on their platforms, thereby solidifying its position in the market.
The DOJ is considering both structural and behavioral changes to prevent Google from exploiting its other products—such as Chrome, Play, and Android—to further its dominance in search. Potential measures include prohibiting Google from using these platforms to promote its search engine over rivals and limiting revenue-sharing agreements that reinforce its position. Assistant Attorney General Jonathan Kanter, who is leading the antitrust efforts, argued that preventing Google’s control over future distribution is just as important as unwinding past behavior.
Google has already indicated it will appeal the ruling and pushed back against the proposed remedies, calling them “radical” and warning of unintended consequences. Lee-Anne Mulholland, Google’s vice president of global affairs, said that splitting off products like Chrome or Android could “change their business models, raise the cost of devices, and undermine competition with Apple’s ecosystem.” Google argued that its search and advertising market shares are still subject to competition, citing TikTok and Amazon as emerging players in the search space.
While the DOJ’s suggestions are yet to be finalized, any action taken would mark a significant escalation in efforts to regulate big tech. Legal experts have noted that while a full breakup may be unlikely, measures to curb exclusive agreements and open up the market to competition could still have a profound impact on the industry’s landscape.
Sources include: BBC, CBS
AI Pioneer Geoffrey Hinton wins Nobel Prize, takes a jab at OpenAI’s CEO
Geoffrey Hinton, often called the “godfather of AI,” made headlines again, not just for winning the 2024 Nobel Prize in Physics but for using the opportunity to take a pointed jab at Sam Altman, CEO of OpenAI. Speaking in a short, impromptu press briefing after the award was announced, Hinton expressed his pride in a former student, Ilya Sutskever, for orchestrating Altman’s firing from OpenAI in a boardroom coup last year. Though the firing was quickly reversed, the episode has left lasting ripples in the AI community.
Hinton, now 76, was humble about the accolade and visibly surprised, running on little sleep after the sudden announcement. He took a moment to thank his mentors, collaborators, and students, crediting them with his achievements in AI. Hinton’s groundbreaking research, notably the creation of the 2012 “AlexNet” with Ilya Sutskever and Alex Krizhevsky, is often referred to as the Big Bang of artificial intelligence for its role in image recognition breakthroughs.
Reflecting on the coup at OpenAI, Hinton explained that his disapproval of Altman stemmed from the CEO’s focus on profit over safety. Altman has faced criticism from former board members and staff for allegedly sidelining OpenAI’s safety team and pushing to shed its nonprofit status in favor of rapid commercialization. “Over time, it turned out that Sam Altman was much less concerned with safety than with profits, and I think that’s unfortunate,” Hinton stated.
Hinton used the platform not only to celebrate the success of his peers but also to underscore the critical need for more research into AI safety. He voiced concerns about the difficulty scientists already face in predicting the behavior of advanced AI systems, cautioning that once these models surpass human intelligence, controlling them may become impossible. “When we get things more intelligent than ourselves, no one really knows whether we’re going to be able to control them,” Hinton said, advocating for urgent efforts to align AI development with human safety.
And that’s our show for today.
And thanks to you for tuning in.
Reach me at editorial@technewsday.ca
I’m your host Jim Love, have a Thrilling Thursday
