Klarna Saves $10 Million in Marketing Costs Through Innovative GenAI Use

Welcome to Towards AGI, your premier newsletter dedicated to the world of Artificial Intelligence. Our mission is to guide you through the evolving realm of AI with a specific focus on Generative AI. Each issue is designed to enrich your understanding and spark your curiosity about the advancements and challenges shaping the future of AI.

Whether you're deeply embedded in the AI industry or just beginning to explore its vast potential, "Towards AGI" is crafted to provide you with comprehensive insights and discussions on the most pertinent topics. From groundbreaking research to ethical considerations, our newsletter is here to keep you at the forefront of AI innovation. Join our community of AI professionals, hobbyists, and academics as we pursue the ambitious path towards Artificial General Intelligence. Let’s embark on this journey together, exploring the rich landscape of AI through expert analysis, exclusive content, and engaging discussions.

TheGen.AI News

Klarna Saves $10 Million in Marketing Costs Through Innovative GenAI Use

Klarna, a fintech company and early adopter of generative AI (GenAI), announced on Tuesday that it uses AI for a variety of functions including conducting marketing campaigns and creating images, achieving an annual cost reduction of about $10 million. The firm has decreased its sales and marketing expenses by 11% in the first quarter, attributing 37% of these savings to AI. This technology has enabled Klarna to boost the number of campaigns while simultaneously reducing image production costs by $6 million, using GenAI tools like Midjourney, DALL-E, and Firefly.

The company updates its app and website imagery weekly to align with significant retail occasions like Valentine's Day, Mother's Day, and summer sales. According to Klarna's CMO David Sandström, using AI has significantly lowered the costs and time associated with producing custom imagery for these events, eliminating the need for traditional stock photos. Over the first three months of 2024, Klarna generated over 1,000 images with GenAI, shortening the image development time from six weeks to just seven days.

Additionally, Klarna saved a further $4 million by reducing expenditures on external marketing services such as translation, production, and social agencies. The company has also collaborated with OpenAI to deploy an AI assistant for customer service, performing the tasks of 700 full-time agents.

HCLTech Debuts AI Force on Google Cloud, Boosting GenAI Innovation and Efficiency

On Thursday, HCL Technologies (HCLTech) unveiled its genAI platform, HCLTech AI Force, now accessible on Google Cloud's Gemini models. This platform, designed to be system-agnostic, integrates Google's advanced multimodal large language model (LLM), Gemini, to enhance innovation at scale, according to the company's announcement from its headquarters in Noida.

HCLTech AI Force not only facilitates the use of Gemini applications but also ensures stringent security and governance frameworks are in place.

Vijay Guntur, the Chief Technology Officer and Head of Ecosystems at HCLTech, highlighted the challenges enterprises face in scaling genAI projects beyond the proof-of-concept phase, often due to high costs. He stated that the HCLTech AI Force platform allows for quicker market entry and increased efficiencies throughout the product engineering lifecycle.

The company also offers industry-specific solutions utilizing Gemini models. These include intelligent asset tracking for manufacturing environments to optimize material movement, and a comprehensive Production Lifecycle Management (PLM) system. Additionally, solutions aimed at mobile providers enhance network automation and support operations aimed at reducing energy use towards net-zero goals.

With its expertise in full-stack application development and engineering, coupled with extensive AI experience, HCLTech is dedicated to helping clients maximize the benefits of gen for applications ranging from chip development to cloud-based services and business process enhancement.

OpenAI CTO Mira Murati Discusses GenAI's Role as Collaborative Tools in Business

At the Asia Tech X Singapore conference, OpenAI's Chief Technology Officer Mira Murati discussed the evolving role of generative AI (GenAI) in the workforce, noting its swift adoption as a collaborative tool in various workplace settings. According to a Bloomberg report, Murati emphasized how GenAI, especially with the introduction of OpenAI’s new model, GPT-4o, is enhancing capabilities in text generation, presentations, and video creation. She also mentioned its growing use in coding, writing, and administrative tasks.

Murati suggested that the economic implications of GenAI are just beginning to manifest. However, a PYMNTS Intelligence study reveals that while almost all enterprise firms employ GenAI in some capacity, many CFOs have not yet observed a significant return on investment. Karen Webster of PYMNTS highlighted that these firms are primarily exploring GenAI through simple, low-risk applications, which currently dominate their use of the technology.

Despite widespread adoption, only about 30% of firms are leveraging GenAI strategically, such as for process monitoring or creative content generation, according to the PYMNTS Intelligence's 2024 C(AI)O Report.

The new GPT-4o model, launched on May 13, promises enhanced speed and improved performance across text, vision, and audio applications. This launch is part of a broader competitive push by tech companies and startups to advance AI technologies.

TheOpen.AI News

Global Gen AI Community Advocates for Open Source Artificial General Intelligence

The 2024 GenAI Summit recently commenced in San Francisco, where Sri Ambati, co-founder and CEO of H2O.ai, advocated for artificial general intelligence (AGI) to be available as open-source technology, equating it to "generational wealth" like music and mathematics. He emphasized that making AGI open source would maximize its freedom and innovative potential.

H2O.ai, Ambati's firm, specializes in creating open-source machine learning and predictive analytics tools aimed at making AI accessible to all by simplifying user interaction and promoting collaborative development. The company also focuses on aspects such as Automated Machine Learning (AutoML), scalability, and performance.

The company introduced its latest model, H2O-Danube2-1.8B, under an Apache 2.0 license, designed for fine-tuning on specific datasets and efficient in both training and inference, suitable for use on edge devices.

Ambati highlighted the integration of generative and predictive AI as crucial for achieving significant productivity improvements and better outcomes, describing it as essential for gaining a holistic view rather than just isolated insights.

The summit, organized by GPT Dao for the second consecutive year, attracted roughly 10,000 attendees and 300 exhibitors, including major tech companies like Microsoft, IBM, and Amazon. GPT Dao, based in Palo Alto and closely associated with Stanford University and UC Berkeley, was established by a local group of students, professors, and entrepreneurs. It focuses on fostering the development of open-source AI technologies and provides various services such as Web3 and AI project incubation, AI education, and a platform for community governance and decentralized crowdfunding.

Meta Embraces Open Source AI, Transforming Industry Perceptions

When Meta's CEO Mark Zuckerberg announced the release of an AI system last year, Jeffrey Emanuel, a passionate AI enthusiast and part-time hacker, was initially skeptical. Emanuel had experience with proprietary AI models like those from OpenAI, which were not open for modification. His concerns grew when Zuckerberg initially introduced Meta’s AI to a select group of academics, fearing the technology would be restricted to a limited group.

However, Zuckerberg's subsequent decision to make the updated AI system open source, allowing free modification and redistribution, changed Emanuel's perspective. As the founder of blockchain startup Pastel Network, Emanuel praised the move, contrasting Meta’s approach with the more guarded strategies of companies like Google, OpenAI, and Microsoft.

Zuckerberg has thus positioned himself as a major advocate for the open-source AI model, diverging from other tech giants who prefer to keep their AI technologies under wraps, citing safety concerns. His stance has not only popularized him within developer communities but also led to Meta’s AI being embraced for a wide range of applications. Since its release, Meta’s open-source AI model, LLaMA 2, and its successor, LLaMA 3, have seen significant adoption and usage, from medical applications to creating digital assistants.

The popularity of Meta’s open-source AI has sparked discussions in Silicon Valley, with some referring to this phase as a "Zuckaissance." Patrick Collison, CEO of Stripe and a member of a Meta advisory group, remarked that the open sourcing of LLaMA might be the most well-received tech initiative by Facebook within the tech community to date.

TheClosed.AI News

OpenAI Disrupts Covert Operations by Chinese, Russian, and Israeli Entities

OpenAI, the company behind ChatGPT, recently uncovered and halted several covert influence operations from countries including Russia, China, Israel, and Iran. These operations involved the manipulation of public opinion and political outcomes through undisclosed and deceptive methods.

The company identified five specific campaigns that leveraged its generative AI models to create sophisticated text and imagery, distributing them across various social media platforms like Telegram, X, and Instagram. The content, often containing fewer linguistic errors than human-generated materials, was part of efforts to influence discussions and perceptions subtly.

OpenAI terminated the associated accounts from two Russian campaigns named Bad Grammer and Doppelganger; a Chinese effort called Spamouflage; an Iranian initiative known as International Union of Virtual Media; and an Israeli operation termed Zero Zeno.

According to OpenAI, these campaigns targeted a range of geopolitical issues. The Russian operations focused on disseminating negative narratives about Ukraine, the U.S., NATO, and the EU. Spamouflage attacked Beijing’s critics and also highlighted historical injustices against Native Americans. The Iranian network produced content critical of the U.S. and Israel, while Zero Zeno targeted a UN agency for Palestinian refugees and extremist elements in Canada.

Despite these extensive efforts, OpenAI reported that these operations did not significantly increase audience engagement or reach through its platform. The company emphasized its commitment to developing AI responsibly, ensuring safety in its models, and actively working to prevent misuse.

This revelation comes amid a broader discussion about the risks of AI in spreading disinformation, especially highlighted by recent election-related incidents in the U.S. and Pakistan where AI was used in attempts to influence voter behavior.

OpenAI Discloses Misuse by State-Backed Groups for Disinformation Campaigns

OpenAI, the organization behind ChatGPT, reported on Thursday that it had disrupted five covert operations that misused its AI technology for deceptive activities over the last three months. The company revealed in a blog post that these campaigns originated from Russia, China, Iran, and a private Israeli company.

The offenders exploited OpenAI's advanced language models for a variety of tasks, including generating social media comments, articles, profiles, and debugging codes for bots and websites. Sam Altman, CEO of OpenAI, noted that these operations failed to significantly boost audience engagement or reach.

Concerns are heightened as OpenAI, known for products like ChatGHP or the image generator Dall-E, faces scrutiny. These tools can rapidly produce misleading content, which is particularly worrying with major elections approaching globally. Countries like Russia, China, and Iran are notorious for using social media to inflame pre-election tensions.

One of the thwarted operations, known as "Bad Grammar," was an undisclosed Russian campaign targeting the Ukraine, Moldova, the Baltics, and the United States, using OpenAI’s models to create succinct political comments on Telegram. Another, the "Doppelganger" operation, utilized AI to produce comments in multiple languages on platforms like X.

Additionally, the "Spamouflage" campaign from China misused the models for social media analysis and multilingual text generation. The "International Union of Virtual Media," an Iranian initiative, was interrupted in its efforts to produce content for government-linked websites. OpenAI also stopped STOIC, an Israeli commercial firm, from using its AI to generate content on platforms like Instagram, Facebook, and Twitter, a campaign which Meta had flagged earlier.

These operations were active across various social platforms including Twitter, Telegram, Facebook, and Medium, yet they failed to captivate a large audience. OpenAI attributes its success in thwarting these misuse efforts to effective collaboration, intelligence sharing, and the built-in safeguards in its AI models. The company's report emphasized the current AI trends of generating large volumes of text and images with fewer errors, blending AI with traditional content, and simulating engagements through AI-generated responses.

OpenAI's ChatGPT Set for Integration into Apple’s iPhone Software

Apple and OpenAI have completed a significant agreement to integrate OpenAI's conversational AI into Apple’s software, potentially valued at billions of dollars. The deal will incorporate OpenAI’s AI products into Apple’s ecosystem, as stated by an anonymous source. 

Apple, which has been perceived as trailing behind other major tech companies in AI development, chose OpenAI over Google for integrating AI chatbots into iOS 18. This decision was reached after discussions with both companies, and an official announcement is anticipated at Apple's upcoming WWDC event, scheduled for June 10-14. 

Despite the strategic partnership, there has been some internal resistance at Apple. John Giannandrea, Apple’s AI/ML chief, expressed skepticism about the value of adding another AI chatbot to the market, questioning the necessity of developing AI-related tools at Apple.

The collaboration aligns with OpenAI’s strategic shift towards commercializing its products to drive revenue, moving beyond purely research-focused initiatives. This shift has sparked discussions, including a recent meeting between Microsoft CEO Satya Nadella and OpenAI’s Sam Altman, where concerns about the deal's impact on Microsoft's existing partnership with OpenAI were discussed.

Furthermore, there is speculation that OpenAI, currently a non-profit, may restructure itself as a for-profit or a benefit corporation (B-Corp), which would allow it to pursue broader goals than just profitability and provide legal protection against shareholder lawsuits when acting beyond mere profit motives.

Join the movement - If you wish to contribute thought leadership, then please fill the form below. One of our team members will be back in touch with you shortly.

Form link

Keep reading

In our quest to explore the dynamic and rapidly evolving field of Artificial Intelligence, this newsletter is your go-to source for the latest developments, breakthroughs, and discussions on Generative AI. Each edition brings you the most compelling news and insights from the forefront of Generative AI (GenAI), featuring cutting-edge research, transformative technologies, and the pioneering work of industry leaders.

Highlights from GenAI, OpenAI and ClosedAI: Dive into the latest projects and innovations from the leading organisations behind some of the most advanced AI models in open-source, closed-sourced AI.

Stay Informed and Engaged: Whether you're a researcher, developer, entrepreneur, or enthusiast, "Towards AGI" aims to keep you informed and inspired. From technical deep-dives to ethical debates, our newsletter addresses the multifaceted aspects of AI development and its implications on society and industry.

Join us on this exciting journey as we navigate the complex landscape of artificial intelligence, moving steadily towards the realisation of AGI. Stay tuned for exclusive interviews, expert opinions, and much more!