Generative AI made Another Big Leap Forward for Business This Week

Our team just got back from the first developer conference for OpenAI - the firm behind ChatGPT — and what I saw took my breath away. The progress in technology innovation, the fall in prices, and the increase in business relevance are all exciting. They’re also a sign of what is to come in the broader GenAI market. I am expecting the pace of innovation from all the platforms across the GenAI market to continue: expanding the ways it can be used as well as reducing the cost and increasing performance for enterprise workloads.  It’s the start of a new era for GenAI.

Whether you’re a CEO considering how GenAI may transform your industry, a COO driving efficiency and speed, a business leader looking for new value, or a CIO working on innovation and production use of AI, the new developments could be game changers. These changes can expand the envelope for what businesses can do with GenAI. Here are my top takeaways.

1. Costs are dropping

Prices for tokens (the units of text or code that GenAI uses to understand what we want and to respond) are falling significantly. OpenAI and MIcrosoft are reducing the cost by two-thirds. This is not the first price drop, and it won’t be the last. The major GenAI vendors have been consistently reducing the cost of their offerings. But this latest drop is significant — and I expect other vendors will likely follow suit. You may now be able to integrate the most advanced AI functionalities at scale with easier ROI thresholds. 

2. GenAI’s ability to handle complex tasks is increasing

Several enhancements were introduced with GPT-4 Turbo, OpenAI’s newest model, enabling GenAI to take on much more complex tasks. It opens up the potential for new use cases exponentially. For starters, the model’s “context window” has been increased significantly (128,000 tokens, compared with the prior model’s maximum of 32,000.) This increase means that the model can “read” more complex documents more quickly and produce more relevant, reliable outputs. It is now more capable, for example, of generating complex legal document analysis, long-form content, and detailed technical assistance. The model also offers a “seed parameter” to enable more consistent results — which are critical for governance and overall performance.

3. A key risk is being addressed — helping manage IP liability

As companies seek to manage the potential risks of generative AI, one is often top of mind: the legal risk of inadvertently using copyrighted material. OpenAI is now matching what several other major GenAI vendors (such as Microsoft, Google and Adobe) have also announced: it is offering to pay its customers’ costs should they face legal claims around copyright infringement. Companies will still need rigorous compliance and governance, but GenAI vendors are helping to address one key risk.

4. Integration is becoming even easier

API advances are making it more seamless to tap into other capabilities and do more with GenAI. For example, GPT-4 Turbo allows API access to OpenAI’s image generation tool (Dall-E 3) and its Advanced Data Analysis tool. You can now more easily input vast amounts of images or data, so ChatGPT can produce customized graphics or granular data analysis. This change could enhance nearly everything that uses data or imagery, such as market research, product design, and the customer experience. At the same time, GPT-4 Turbo can also now “call” multiple functions in a single API message. This means that it can perform multiple tasks with lower latency (increased speed) and less computational overhead (lower hardware costs). 

5. You can customize GenAI more easily

OpenAI has made it easier for you to build your own versions (“GPTs”) of ChatGPT, trained on domain-specific or proprietary data. You can create agents for specific tasks (such as legal review or customer service) or specific domains, such as HR or your main lines of business. The result of this customization can be greater relevance and reliability, enabling your people to apply the technology in more tailored ways as they reshape how their work is done.

What it means for the future of business 

The advances we’re seeing are far bigger than any single company. They are part of the larger innovation cycle we have been in for the last 2 years and which is only accelerating. Altogether, GenAI can now do far more, and do it more affordably, than it could a month ago. And we’re just getting started.

The most important thing to know? It’s important to think big: Consider how to implement GenAI at scale and achieve rapid ROI. Our method starts with trust, data governance and security. It deploys an AI factory for fast, repeatable results. It uses a scalable, modular approach to building use cases.  It applies responsible AI practices and emphasizes workforce transformation: Employees should understand both how to use GenAI responsibly, and how to reimagine their roles. With a human-led, tech-powered approach, GenAI can enhance nearly every aspect of most knowledge workers’ jobs. 

No matter who your cloud or GenAI provider may be, the message from this week’s developer conference is clear: GenAI is ready for business, at scale.