The headlines scream of a revolution – Artificial intelligence (AI) is transforming lives, boosting businesses, and even displacing workers. But amidst this information overload, a crucial question remains unanswered: Which AI reigns supreme?
To navigate this complex landscape, I delved into the experiences of executives, engineers, and researchers at the forefront of applying AI to real-world problems. Their insights, both surprising and revealing, offer a roadmap for businesses and a glimpse into the future of this burgeoning field.
Today's most powerful AIs are cloud-based, readily accessible for testing. However, this accessibility comes at a cost – limited user control over their behavior. Mark Daley, Chief AI Officer at Western University, emphasizes: "It's more akin to hiring an employee than buying software." Each AI, from Google's AI to OpenAI's and Anthropic's offerings, possesses unique strengths and weaknesses. The key to success lies in experimentation and finding the best fit for the user and the specific task.
OpenAI's models continue to be the benchmark against which others are measured. Companies like Nomad Data leverage OpenAI's capabilities to digest, summarize, and search through massive document libraries. Brad Schneider, Nomad Data's CEO, highlights OpenAI's edge in document processing tasks, attributing it to superior accuracy compared to competitors like Anthropic's Claude and Google's Gemini, both prone to "hallucinations" (generating inaccurate information).
Accuracy reigns supreme, but speed and cost are equally important considerations, notes Eric Olson, CEO of Consensus, a scientific search engine. Their platform relies on near-instantaneous summaries of research papers generated by AI. For Olson, OpenAI's GPT-4 "turbo" stands out, delivering responses within 1.5 seconds, compared to the double wait time for the standard GPT-4. However, this blazing speed comes at a premium – OpenAI and its competitors charge businesses per "token" (essentially, per word) processed.
The field of generative AI is experiencing an evolution reminiscent of the early internet boom. Anthropic's recent release of a model rivaling OpenAI's, despite having a smaller and younger team, suggests a potential commoditization of large language models. The future could see fierce competition for the fastest and most affordable solution. While this benefits companies by boosting employee productivity at a fraction of the human cost, it raises unsettling questions about the future of white-collar jobs.
As the AI arms race intensifies, businesses must stay informed to harness the power of these evolving technologies. Experimentation with various AI models is key to finding the optimal fit for specific needs. The coming years will see a dramatic shift in the AI landscape, with potential benefits for businesses alongside potential challenges for the workforce. By staying engaged and proactive, companies can leverage this revolution to unlock new levels of efficiency and innovation.
