Flag

We stand with Ukraine and our team members from Ukraine. Here are ways you can help

Home ›› Artificial Intelligence ›› Open AI Dev Day: Great Technology, Light on UX and Business Value

Open AI Dev Day: Great Technology, Light on UX and Business Value

by Rich Weborg
5 min read
Share this post on
Tweet
Share
Post
Share
Email
Print

Save

As someone who’s been working with the various versions of GPT and other large language models for a number of years, I was excited when ChatGPT supercharged the adoption of (and interest in) conversational AI. By putting a chat interface on their pre-trained language model, OpenAI played a major role in ushering in an accelerated and exciting era in technology. In a rapidly growing field of LLMs, their GPT models remain near the top. Attending the recent OpenAI Dev Day, I noticed that their team—including CEO Sam Altman—did a great job presenting their platform, but their mantra was “we can’t wait to see what you guys create.” 

This felt like code for, “we know this is important, but we’re not sure how, or for what.” Sadly, this theme endured through most of what was presented, I didn’t see a whole lot of holistic design in how their virtual agents (they’re calling them GPTs) will integrate into company architectures and our daily lives. The focus on UX was lacking, and perhaps by extension, so was a sense of how businesses will generate value with generative AI. It left me feeling that the industry as a whole does not yet know how to drive real business value out of the technology that OpenAI and other generative AI platforms provide. Here are some things that stood out to me about the conference and how these technologies were presented. 

What About Brand Alignment?

There are massive opportunities for businesses that can align AI with their brand identity, but I didn’t hear this addressed. Speakers partially hit on UX, showing how to customize responses (e.g. friendly, serious, detailed responses, short replies) but seemed to completely ignore the value of matching an AI to the deeper elements of a company’s brand. For example, a company like Peloton might want experiences with generative AI to model the ways their instructors communicate to riders. This could include combination tones, ranging from chipper and supportive (i.e. “You’ve got this. I know you can do it!) to challenging and blunt (e.g. “If you can’t show up for yourself, how can you show up for anyone else?”). This kind of alignment also demands lexicography or brand-approved terms and usages.

The conference lacked discussions on how to control AI responses effectively. I was expecting demonstrated mechanisms for refining AI interactions to ensure they align with a brand’s messaging and guidelines, but I didn’t hear anything about how to control what the AI can respond to or to what degree. They also missed an opportunity to provide examples of customized avatars, names, and conversational tones that could resonate with a company’s image, creating a more cohesive and engaging user experience.

How Will it Work on Different Platforms?

In terms of harnessing the power to create next-level process automation, OpenAI could have explored various platforms where their GPTs can be deployed, such as embedded in websites, in mobile apps, and used for interactive voice response (IVR) systems. Given the number of businesses across industries that will be looking to integrate generative AI, flexibility and adaptability across channels seems paramount. Conversational AI can enhance UX across diverse touchpoints, but only after these considerations have been addressed.

 What About Handling Unanswered Questions?

There was no acknowledgment of what happens when AI cannot provide satisfactory responses. Businesses need strategies for gracefully handling these situations, including options like politely declining to answer, transitioning the conversation to a different topic, or (in a business setting) connecting the user with a human in the loop. Properly designed experiences can’t leave users frustrated when faced with AI limitations. A user is less likely to re-use an agent that can not get a question answered or a task completed. Great exception handling is the bedrock of great UX.

Will There Be Human-AI Interaction?

There are so many use cases where interactions with AI can contribute to rewarding UX. These require a strategy for integrating AI seamlessly into customer service processes. This includes scenarios where AI is collaborating with human agents (the human-in-the-loop I referenced above). There’s a ton of value in creating a harmonious balance between AI and human interactions and I didn’t hear anything about it at Dev Day. This was a bit surprising given that Gartner recently predicted that by 2025, generative AI will be a workforce partner for 90% of companies worldwide [1]. Interactions between humans and machines shouldn’t be left to design themselves. These relationships will require careful calibration and ongoing collaboration between humans and machines.

Interesting and somewhat useful on their own, many of these apps could become massively rewarding if they were part of business-wide use cases.

Where Are the Real Business Use Cases?

The conference featured many intriguing examples of conversational AI in action—talking like a pirate, controlling lights, explaining board games—but they were ultimately rather frivolous. These are cool applications that seemed nearly impossible only a few years ago, but companies are looking for practical business use cases. OpenAI could have highlighted some of them. Tasks like appointment scheduling, feedback collection, real-time invoice generation, and customer support are ripe for automation using generative AI. According to a recent CNBC report, almost half of top technology officers across industries are putting AI as their top budget item over the next year, while nearly two-thirds indicated that their AI investments are accelerating. This was a staggering missed opportunity.

My Takeaways

Despite these shortcomings, the conference was thrilling. For technologists, it’s hard to imagine a more exciting time to be alive. For those of us who have been working in conversational AI for years, we finally have a way to make real human-machine interactions meaningful. Tools like OpenAI’s platform make it possible to avoid piecing applications together with various NLU tools and strict dialog-driven approaches.

However, this industry is still in its inception. For as fast-evolving as generative AI is, we can’t forget that it is just a tool, like a hammer, and not everything is a nail. It’s mission-critical to focus on real business outcomes and the UX that drives them.

We’ve all seen generative AI write poetry. It’s time to see it start generating real business value.

  1. Gartner, Gartner IT Symposium 2023 Presentation, “We Shape AI – AI Shapes Us”, Mary Mesaglio and Don Scheibenreif, Oct. 16-19, 2023
post authorRich Weborg

Rich Weborg
Rich Weborg, is a founder at OneReach.ai with 30+ years of experience in conversational design, solution architecture, and technology implementations. He’s passionate about the potential of technology to foster meaningful relationships between humans, and between humans and machines.

Tweet
Share
Post
Share
Email
Print
Ideas In Brief
  • The article provides constructive feedback on OpenAI’s Dev Day, emphasizing opportunities for improvement in brand alignment, platform adaptability, and practical business use cases, while recognizing the exciting potential of conversational AI.

Related Articles

Is true consciousness in computers a possibility, or merely a fantasy? The article delves into the philosophical and scientific debates surrounding the nature of consciousness and its potential in AI. Explore why modern neuroscience and AI fall short of creating genuine awareness, the limits of current technology, and the profound philosophical questions that challenge our understanding of mind and machine. Discover why the pursuit of conscious machines might be more about myth than reality.

Article by Peter D'Autry
Why Computers Can’t Be Conscious
  • The article examines why computers, despite advancements, cannot achieve consciousness like humans. It challenges the assumption that mimicking human behavior equates to genuine consciousness.
  • It critiques the reductionist approach of equating neural activity with consciousness and argues that the “hard problem” of consciousness remains unsolved. The piece also discusses the limitations of both neuroscience and AI in addressing this problem.
  • The article disputes the notion that increasing complexity in AI will lead to consciousness, highlighting that understanding and experience cannot be solely derived from computational processes.
  • It emphasizes the importance of physical interaction and the lived experience in consciousness, arguing that AI lacks the embodied context necessary for genuine understanding and consciousness.
Share:Why Computers Can’t Be Conscious
18 min read

AI is transforming financial inclusion for rural entrepreneurs by analyzing alternative data and automating community lending. Learn how these advancements open new doors for the unbanked and empower local businesses.

Article by Thasya Ingriany
AI for the Unbanked: How Technology Can Empower Rural Entrepreneurs
  • The article explores how AI can enhance financial systems for the unbanked by using alternative data to create accessible, user-friendly credit profiles for rural entrepreneurs.
  • It analyzes how AI can automate group lending practices, improve financial inclusion, and support rural entrepreneurs by strengthening community-driven financial networks like “gotong royong”.
Share:AI for the Unbanked: How Technology Can Empower Rural Entrepreneurs
5 min read

Curious about the future of AI? Discover how OpenAI’s “Strawberry” could transform LLMs with advanced reasoning and planning, tackling current limitations and bringing us closer to AGI. Find out how this breakthrough might redefine AI accuracy and reliability.

Article by Andrew Best
Why OpenAI’s “Strawberry” Is a Game Changer
  • The article explores how OpenAI’s “Strawberry” aims to enhance LLMs with advanced reasoning, overcoming limitations like simple errors and bringing us closer to AGI.
  • It investigates how OpenAI’s “Strawberry” might transform AI with its ability to perform in-depth research and validation, improving the reliability of AI responses.
Share:Why OpenAI’s “Strawberry” Is a Game Changer
3 min read

Tell us about you. Enroll in the course.

    This website uses cookies to ensure you get the best experience on our website. Check our privacy policy and