{"id":70182,"date":"2024-04-15T05:00:00","date_gmt":"2024-04-15T12:00:00","guid":{"rendered":"https:\/\/www.salesforce.com\/?p=70182"},"modified":"2024-04-18T13:27:10","modified_gmt":"2024-04-18T12:27:10","slug":"large-language-models-scale","status":"publish","type":"post","link":"https:\/\/www.salesforce.com\/uk\/news\/stories\/large-language-models-scale\/","title":{"rendered":"AI Balancing Act: How Companies Can Scale LLMS to Improve Performance, Cut Costs, and Help the Environment"},"content":{"rendered":"\n

A media feeding frenzy over the exploding scale<\/a> and increasing costs<\/a> of artificial intelligence (AI) has created a divide between the reality of the technology and its perception among decision makers, explained Silvio Savarese<\/a>, Chief Scientist of Salesforce AI Research in a recent interview.<\/p>\n\n\n\n

Whether these concerns come from worries about the bottom line, the impact of AI on the environment, or even basic questions of fairness and access, Savarese believes that changing these misconceptions requires an understanding of when scale is necessary to deliver high-quality AI outputs \u2014 and when it isn’t.<\/p>\n\n\n\n

In the interview, Savarese also shared why the environmental and financial price tag of AI doesn\u2019t have to be as head-spinning as the headlines might suggest, and how understanding the different scales and performance of AI models can help any business responsibly harness this transformative technology to boost productivity, build deeper relationships with customers, and enhance daily workflows and processes.<\/p>\n\n\n\n

Q. Today\u2019s well-known Large Language Models (LLMs) are getting a lot of negative attention for the compute power it requires to run them \u2014 both in terms of cost to operate as well as environmental impact. Are models this large necessary for businesses to tap into the power of generative AI? <\/strong><\/h3>\n\n\n\n

Rather than asking if the scale of today\u2019s LLMs is necessary, let\u2019s ask what it\u2019s necessary for<\/em>.<\/p>\n\n\n\n

The scale and size of an AI deployment isn’t inherently advantageous. Rather, when implementing AI, there\u2019s a range of possibilities and trade-offs that should be explored. Remember, the ChatGPTs of the world are designed to do more or less everything, and that makes them very different from most enterprise applications. They can help with homework, suggest holiday recipes, and even reimagine La boh\u00e8me’s libretto as Socratic dialogues. It\u2019s a great party trick\u200c \u2014 \u200calbeit an expensive one\u200c. Training Open AI\u2019s ChatGPT 4 cost more than $100 million<\/a>. \u200cBut that isn\u2019t what enterprises are using AI for.<\/p>\n\n\n\n

\n

The scale and size of an AI deployment isn’t inherently advantageous. Rather, when implementing AI, there\u2019s a range of possibilities and trade-offs that should be explored.”<\/p>\nSilvio Savarese<\/a>, Chief Scientist of Salesforce AI Research<\/cite><\/blockquote>\n\n\n\n

There\u2019s also the environmental impact of these large LLMs. The hypothetical long-term benefits of AI in combating climate change in areas such as monitoring emissions and optimizing the transportation of goods are significant, with the potential to reduce global emissions 5 to 10% by 2030<\/a>. However, the utilization of LLMs, while groundbreaking in their capabilities, requires enormous computing resources<\/a>, exacerbating pressing concerns such as the release of greenhouse gasses, the depletion of water resources, and the extraction of raw materials along the supply chain. Given the urgency of the climate crisis and the imperative to combat planet-warming emissions, it’s paramount that the development and implementation of AI technologies doesn’t surpass the capacity of our planet’s resources.<\/p>\n\n\n\n

In contrast to LLMs like ChatGPT or Anthropic\u2019s Claude, an AI model like our own CodeGen 2.5<\/a>, has a limited set of tasks \u2014 \u200chelping developers write, understand, and debug code faster. Despite its deliberately small scale, its performance is on a par with models literally twice its size<\/a>, boasting remarkable efficiency without compromising on utility. So even as it helps developers work faster, it also reduces costs, latency, and, crucially, environmental impact<\/a> compared to larger-scale LLMs. <\/p>\n\n\n\n

Businesses should not be asking whether<\/em> they need scale, but how they want to apply<\/em> that scale. Depending on the task, the answer may vary wildly\u200c \u2014 \u200cand bigger is most certainly not always better.<\/p>\n\n\n\n

Q. Okay, but large models still outperform smaller ones, right?<\/strong><\/h3>\n\n\n\n

Believe it or not, even this isn\u2019t a clear-cut answer. Large models do<\/em> generally outperform their smaller counterparts when it comes to flexibility. But therein lies the nuance that is so often left out of conversations around LLMs: as tasks become more narrow, more well-defined, and more unique to a specific organization or domain\u200c \u2014 \u200cexactly what enterprise AI is all about\u200c \u2014 \u200cit\u2019s possible to do more with less. <\/p>\n\n\n\n

In other words, most models aren\u2019t meant to be everything to everyone, which frees up enterprises to focus on their needs while saving huge amounts of resources in the process.<\/p>\n\n\n\n