The Next Step Forward in AI will come from Consumer Businesses

OpenAI’s ChatGPT has grown faster than any consumer app in history, reaching ~100M weekly active users globally in the first year since launch. Two initial observations:

First, despite the wide availability of Generative AI tools, I am not directly using ChatGPT or one of its competitors at work or at home very much – and I suspect I’m not the only one. Luckily, this technology transition is not like the smartphone revolution where the consumer needed to change their behavior (ie. buy a smartphone) to get in on the fun. In fact, almost all the consumer businesses I use daily are using Generative AI tools to improve their products and services – according to OpenAI, more than 92% of Fortune 500 companies use the platform. So I may be getting more advantage from AI than I know.

Chris Dixon famously posited that “the reason big new things sneak by incumbents is that the next big thing always starts out being dismissed as a ‘toy’.” Then technologies get better at a faster rate than our needs increase, and presto the startups that bet on the new tech have overcome the incumbents. That’s not happening here. Certainly OpenAI (and others) beat Google and MSFT’s to the Generative AI punch, but now that the core models exist, no company is skipping this change. And if anything, the advantage has gone to the incumbents with distribution over the startups with innovation this time (See Alex Rampell, Distribution vs. Innovation).

Second, it feels like the various underlying LLMs available to me for text and video are quickly becoming good enough at the basic things that I need them for that they are interchangeable if not commoditized. While I don’t doubt that there will be continued improvements, I do doubt that the improvements will be limited to one LLM. So consumers and businesses need not fear picking the wrong LLM out of the 4-5 leaders today. Just pick one.

‍

Next Steps for Generative AI‍‍‍

So where will we get the most benefit from Generative AI? It’s going to come from the consumer businesses we all use daily.

One of the most common insertion points for them will be in business processes that currently require expensive human to human interaction – whether in the distributed workforce model (eg. healthcare, wealth management) or in centralized models (eg. contact centers for support and sales). Particularly with businesses at scale, the impact Generative AI will have on their workforce can’t be understated.

To that point, AI will certainly replace humans in some jobs, and it will be a massive deflationary force (see past transitions where tech brought down the price of “exclusive” goods like phones and cars). But it will also provide efficiencies/tools that will enable workers to achieve outcomes never before imagined, creating entirely new categories of jobs (see the transition to eCommerce and next day delivery, which eliminated retail job but created jobs for pick, pack and delivery that now make Walmart and Amazon the largest private employers in the US). I think the transition will come to centralized contact center teams first, and will happen in about three phases:

‍

How fast will the transition to AI happen?‍

We are already seeing the transition into the AI-Assisted phase, and many of our customers are using our AI tools (eg. suggested messaging,bots for conversational ai, conversation intelligence and coaching). But it’s harder to predict when we will move into the third phase. Even if the AI is not indistinguishable from a human yet, I am sure we are not far from having every basic support interaction answered by Generative AI. And I am not talking about AGI (Artificial General Intelligence); just AI that surpasses the quality and speed of current interactions. The speed of response 24/7 that Conversational AI can bring is important enough that I would sacrifice a lot of bedside manner to get the answers immediately as long as they are right (or at least more right than the current human team).

And the business impact is immediate. For instance, even in the more complex (than support) sales interactions, DemandFarm showed that when businesses use AI to aid sales processes, they increase leads by 50% and reduce both call times and overall costs by about 60%, and UniqueTimes released a study that 61% of the sales team who have harnessed the power of AI in the sales process have exceeded their revenue goals by 61%.

If you want the most aggressive case, Zack Kass, Former Head of GTM at OpenAI, suggested in my B2C Sales Podcast that AI is already indistinguishable from a human. But he predicted that regulation will slow down the rollout.

Based on the AI tools I’ve used is that Phase 3 “AI Instead of Humans” is still years off. Here’s why:

• AI training data isn’t complete. AI models don’t always include the most relevant and recent content; many models are trained only through 2021, for example, and require companies to train the models on their own data.
• AI outputs are still not reliable. Models sometimes “hallucinate” or exceed their remit. Automated identification and these issues and human oversight is necessary, especially in sensitive industries.
• AI voice technology isn’t yet up to par. Text-based interaction is improving quickly, but the ability to listen and speak is still slow and doesn’t handle cases like interruptions and changes of topics well. 
• AI cannot read human emotion well. Human sales reps are able to understand emotion and express the right response, which AI isn't good at yet. 
• Regulation. The fear of AGI creating Terminator-like outcomes will result in the creation of regulatory frameworks that slow advancement in consumer-facing AI for the public good.
  

How should my consumer business use Generative AI?

At Regal, we are changing the way businesses and people connect online. Our software helps businesses drive more revenue from their sales and support interactions by generating real-time customer intent data to intelligently personalize the calls, texts, emails, and chats between customers and digital businesses. Our advantage over other AI systems is that we have a Unified Customer Profile with a rich understanding of a customer’s real-time first party data online and in interactions with the brand that the AI can use to personalize interactions.

AI is playing a role for Regal’s customers already in a number of ways:

1. Creativity - Generative AI to helps marketers and sales leaders develop new messages
2. Optimization - Machine learning optimizes what the next customer experience is when and with what message to maximize engagement rate and outcome.
3. Customer Interactions - Suggesting responses to agents, and in some cases responding instead of agents. Counter-intuitively, AI Bots may enable every business to offer more of a “human touch” than they do today.
4. Insights and Coaching - Transcription and analysis driven to extract key themes that will drive operational changes to everything from customer service training and sales scripts to manufacturing and delivery operations.

As AI-enabled use cases like these do not require any change on the part of consumers, they will become common with more speed than past transitions like the internet and the smartphone. Regulation may slow down the use of more advanced forms of AI, but even the LLMs available today will enable great improvements in business operations.‍

Learn about these products and services at Regal.io, read more posts at regal.io/blog or email us at hello@regal.io.

‍