Enterprise Artificial Intelligence (AI) has been making waves with the emergence of popular chatbots like ChatGPT and Gemini. However, executives engaged in the buying and selling of AI technology are determined to push the adoption of enterprise AI beyond its nascent stage. Recent developments by top AI providers reflect this commitment, as they unveil business-focused updates and innovative AI models for large language processing.
An enterprise-focused company, Cohere, has introduced a cutting-edge language model (LLM) called R+. Developed for Microsoft Azure and Oracle, this new solution aims to enhance the accuracy and relevance of LLMs while offering support for 10 languages. On the other hand, OpenAI has expanded its capabilities, enabling companies to fine-tune and customize AI models. Additionally, Amazon has further extended its enterprise tools by integrating AI models from Mistral, a French firm.
Nevertheless, a new survey conducted by Writer AI, an enterprise-focused startup that launched its own Palmyra LLM in January, highlights some concerns regarding the development of AI solutions for enterprises. According to the survey, only 17% of the 500 executives polled reported success with their internally built generative AI tools, while 61% encountered accuracy issues. While these findings may seem favorable for a company providing private AI solutions to enterprises, they also contribute to the ongoing debate surrounding whether companies should build their own AI tools or leverage technology from external sources.
Despite these concerns, multiple forecasts indicate significant growth in different aspects of enterprise AI. The advertising industry, in particular, is poised to benefit greatly. A report by Bloomberg Intelligence predicts that ad spend driven by generative AI could grow from $4.6 billion in 2023 to a staggering $206 billion by 2032. Additionally, analysts expect substantial software revenue growth for consumers and e-commerce, with estimates projecting a rise from $995 million in 2023 to $45 billion in 2032.
At an AI event organized by Bloomberg Intelligence, executives responsible for selling AI models and computing power expressed their observations on the growing interest from enterprises. Advances in AI models, coupled with increased computing power, have sparked this newfound interest. However, some experts believe that the adoption of generative AI in the enterprise sector is still in its infancy.
According to Brian Venturo, co-founder and chief strategy officer at CoreWeave, a cloud-based computing provider, the true potential of enterprise adoption has not yet been realized. Venturo remarks, “When I hear people say that enterprise apps are currently limited, I struggle to understand the use case at hand. It feels like we’re just scratching the surface, creating trivial applications without fully harnessing the power of generative AI.”
Neerav Kingsland, Head of Global Accounts at Anthropic, affirms that last year was primarily focused on proof-of-concept trials rather than full-scale deployment by enterprise customers. However, this trend is beginning to shift. Anthropic’s models are being utilized for legal analysis by LexisNexis, financial analysis by hedge funds like Bridge Water and Jane Street Capital, and chat summarization by platforms like Slack.
While some executives remain underwhelmed with the progress of generative AI adoption, others are already seeing the benefits of developing their own AI tools internally. Companies like Mastercard believe that building generative AI tools assists in promoting adoption across global markets and diverse customer segments. During the Bloomberg Intelligence event, Raj Seshadri, President of Data and Services at Mastercard, emphasized the unique capabilities their organization possesses. Seshadri stated, “Working at scale and deploying customized solutions locally has been a significant advantage for us. Many of our global peers collaborate with us because they can leverage our offerings at a lower cost and achieve greater effectiveness.”
FAQs
-
What is generative AI?
Generative AI is a branch of artificial intelligence that involves the creation of artificial data or content, such as text, images, or videos, by an AI model. -
What are language models (LLMs)?
Language models (LLMs) are AI models specifically designed to analyze and understand human language. They can generate text or provide insights based on natural language input. -
How does fine-tuning AI models work?
Fine-tuning AI models involves modifying pre-trained models to improve their performance in specific tasks or contexts. This process allows companies to customize the models according to their unique requirements.
In the pursuit of integrating AI into enterprise operations, companies must strike a delicate balance between protecting sensitive information and exploring the potential of AI models. Navrina Singh, Founder and CEO of Credo AI, an AI governance platform, emphasizes the need for caution. She suggests that companies should test AI systems within controlled environments to mitigate risks associated with unexplored areas such as toxicity, hallucinations, IP leakage, and other emerging challenges. Singh states, “We must prioritize explainability and transparency in these systems. Without a comprehensive understanding of how they operate and respond, we risk facing unknown consequences.”
Another significant hurdle hindering the deployment of enterprise-grade AI is concerns regarding the accuracy and security of data. Writer AI’s survey discovered that 95% of respondents believe additional security measures are necessary, while 94% expressed concerns about data protection.
John Roese, Global CTO at Dell, draws a parallel between large language models and search engines. He suggests that relying solely on public AI models may lead enterprises to lose a substantial competitive advantage by compromising their data. Roese comments, “Public large language models resemble search engines in that they have multiple functionalities. However, enterprises surrender a significant edge when they entrust their data to these public AI models.”
Prompts and Products: AI announcements and other news
- Yahoo has acquired Artifact, an AI news platform founded by Instagram’s co-founders.
- Scott Donaton, former top marketer at Hulu, has joined VersusGame, an AI mobile content startup, as its new CMO.
- Amazon and Hugging Face, an open-source AI platform, are embarking on a North American “roadshow” to engage with developers.
- Intel has announced the fourth cohort for its startup accelerator program, Intel Ignite.
- During an interview with Bloomberg, YouTube’s CEO Neal Mohan suggested that OpenAI’s text-to-video platform, Sora, may have violated the platform’s policies if YouTube content was used for training.
- A report by eMarketer indicates that marketers believe generative AI is not meeting its perceived potential.
Quotes from Humans: AppLovin’s Adam Foroughi on AI and social commerce
AppLovin, a major player in ad-tech, aims to assume a larger role in the realm of social commerce. The company recently announced a $50 million strategic investment in Flip, a startup that facilitates user searches and shopping through an online marketplace featuring influencer reviews and user-generated content. AppLovin’s new ad marketplace for merchants will launch through this partnership with Flip.
Enterprise Artificial Intelligence (AI) continues to gain momentum as companies strive to push its adoption beyond its early stages. Top AI providers like Cohere, OpenAI, and Amazon are unveiling business-focused updates and innovative AI models for large language processing. For example, Cohere has introduced R+, a cutting-edge language model developed for Microsoft Azure and Oracle to enhance accuracy and support multiple languages. OpenAI now allows companies to customize AI models, while Amazon has integrated AI models from Mistral into its enterprise tools.
However, a recent survey by Writer AI raises concerns about the development of AI solutions for enterprises. The survey shows that only 17% of executives reported success with internally built generative AI tools, with 61% experiencing accuracy issues. This highlights the ongoing debate around building AI tools internally versus leveraging technology from external sources.
Despite these concerns, industry forecasts predict significant growth in enterprise AI. The advertising industry, in particular, is expected to benefit greatly, with generative AI-driven ad spend projected to reach $206 billion by 2032. Software revenue for consumers and e-commerce is also expected to experience substantial growth.
While the interest in enterprise AI is growing, some experts believe that the adoption of generative AI is still in its early stages. Executives in the AI industry expressed their observations at an AI event organized by Bloomberg Intelligence. Brian Venturo, co-founder of CoreWeave, believes that the true potential of enterprise adoption has yet to be realized. Neerav Kingsland, Head of Global Accounts at Anthropic, adds that while last year focused on proof-of-concept trials, full-scale deployment by enterprise customers is now on the rise.
Some companies, like Mastercard, believe in the benefits of building their own AI tools internally to promote adoption across global markets and diverse customer segments. Raj Seshadri, President of Data and Services at Mastercard, emphasized their organization’s unique capabilities and cost effectiveness.
However, caution is necessary in integrating AI into enterprise operations. Navrina Singh, Founder and CEO of Credo AI, emphasizes the need for testing AI systems within controlled environments to mitigate risks associated with unexplored areas, such as toxicity and IP leakage.
Data accuracy and security also remain significant concerns hindering the deployment of enterprise-grade AI. The Writer AI survey found that 95% of respondents believe additional security measures are necessary, while 94% expressed concerns about data protection.
John Roese, Global CTO at Dell, cautions against solely relying on public AI models and suggests that enterprises may compromise their data and lose a competitive advantage. He recommends prioritizing private AI models for sensitive data.
In summary, while the AI industry strives to further develop enterprise AI, concerns regarding accuracy, security, and the use of public models remain. However, industry forecasts point to significant growth in various aspects of enterprise AI, and companies are exploring different approaches to realize its full potential.
The source of the article is from the blog toumai.es