Whenever a new technology surfaces, it’s natural for strong emotions to emerge and create polarizing views. Questions around AI replacing humans in their jobs have been around for years, but once people saw what ChatGPT could do, this conversation became slightly more real. We are being told by some that if we don’t jump aboard the AI train now, we will be doomed forever!
Personally, I like to take a more balanced view. Yes, OpenAI’s ChatGPT and Generative AI have been amazing for the world to experience, but it’s still early days – and the best is yet to come! Now I’d like to explore a few of my own predictions for Generative AI in the Salesforce ecosystem.
1. AI Skills Will Become a Requirement for 100% of Roles
Although I’ve started this article by explaining my more ‘balanced’ view, this first prediction may seem a bit bold. Let me explain my logic…
Whilst ChatGPT is impressive, from a user’s point of view, it’s just a search engine on steroids! Large language models (LLMs) are essentially trained using vast amounts of information from the internet. Now, although it’s not explicit, I think it’s safe to say that 100% of Salesforce roles require search engine skills, as the ability to quickly answer questions allows you to be more productive in your role.
But AI is a lot more expansive than search engine skills alone. We will be using AI tools to help us in our own technical Salesforce roles, implementing AI functionality on the Salesforce platform, as well as supporting users who will be utilizing Sales & Service GPT.
Amongst all of these different tools we will need to support and use ourselves, we also need to take into account trust and privacy – a concern that is at the forefront of everyone’s minds. Two recent products all Salesforce professionals will need to get their heads around to meet these requirements are the Salesforce AI Cloud and Prompt Studio.
But before launching into a full-blown panic about your lack of AI skills, just remember how ‘early days’ we are. ChatGPT was only released at the end of November 2022, and Salesforce are only just beginning to announce and release their AI features.
You are also not alone; in a recent survey of 4,000 desk workers, they found that while 54% of workers believe generative AI will advance their career, 62% say they don’t have the skills to effectively and safely use the technology.
So, although AI is coming at us thick and fast, no one in the Salesforce ecosystem is truly an expert yet, and we all have time to level up our skills before this becomes a requirement.
2. AI Will Not Replace Experience
I use ChatGPT as much as the next professional. I’ve found that its creativity and its ability to surface complex, customized information quickly is incredible. But I would never use any of its outputs without verifying them first, and most likely, changing quite a few details.
In my day job I’ve tried using it for writing sales emails, marketing copy, and a few social captions. It’s been fantastic for creating a few ideas, but the final result was me writing for myself and using the ChatGPT output for inspiration, which admittedly did speed up the process. Why didn’t I just use the output that ChatGPT provided? Because, frankly, I thought I could do a better job!
Salesforce professionals can currently use ChatGPT to create validation rules, Apex Triggers, Lightning Components, and a lot more. But would you ever use an output without validating it? Of course not, and to validate it requires experience.
This is why we may be entering a tricky time for entry-level professionals joining the world of Salesforce, or really any tech ecosystem where ChatGPT is used for support. Too much reliance on generative AI at the early stages of your career may provide you with a disadvantage, as the ability to deeply understand the outputs may be lost.
To take an example, one of the most anticipated features for Salesforce professionals is Einstein GPT for Flow, which gives you the ability to create flows using a prompt e.g. “Create a flow to send an email to the opportunity owner when the stage changes to ‘closed won’”.
Whilst I’m sure that many Salesforce professionals will be rejoicing at this feature announcement, the ability to verify the outputs is vitally important, and that only comes with experience being built up outside of using Generative AI.
3. Every Salesforce Professional Will Have an AI Assistant
If we try and think what the world will look like 30 years from now, I don’t think it’s too farfetched to imagine that most homes will have some sort of artificial general intelligence (AGI) – which will be as ubiquitous as our Amazon or Google smart speakers. Perhaps they will cook food for us, monitor our health, or even diagnose why the printer won’t connect!
But AI assistants for Salesforce professionals could be right around the corner. ChatGPT has given us an insight into what this assistant could do for us e.g. create Apex code, validation rules, or even write out the steps to create a flow. But there are a few obvious drawbacks…
OpenAI’s latest LLM (Large Language Model) is GPT-4, which uses 170 trillion parameters and can answer questions on essentially anything, so it’s extremely generic. ChatGPT’s interface also doesn’t help with specific Salesforce questions; there are no hints, or pre-made prompts to help you answer your questions.
Obviously this isn’t what ChatGPT was built for; it was built for everyone. So what happens when people start building purpose-made tools to assist technology professionals, and in particular, Salesforce professionals?
We are seeing this in tools such as Github’s Copilot, and will soon start to see the impact of Einstein GPT for Flow, as well as Einstein GPT for Developers. This will improve developer productivity with an AI chat assistant to generate code and ask questions for languages like Apex, all powered by Salesforce’s proprietary LLM.
Productivity AI tools could soon be available for every desk job on earth. When we start combining a highly trained LLM for a specific use case, mixed with a user interface that supports a specific role with common problems, magic really starts to happen.
4. Salesforce Users Will Be Happier
It’s hard not to get caught up in the excitement of AI and the promise of how disruptive this will be to the entire world. But it’s also sensible to be cautious – after all, if you’re excitable like me, you may have been caught up in the recent waves of the metaverse, NFTs, and crypto, only to be pretty disappointed.
But AI is a bit different than the latest fad. ChatGPT has shown the world what is possible with the first widely accessible large language model and simple interface, with most people being blown away with its outputs. But of course, we are only at V1 of what AI can accomplish for us.
V2 will come in the form of existing platforms integrating with large language models provided by companies such as OpenAI, delivering further insights based on the extra context provided. Salesforce has recently announced 16 products that will be rolling out during the “Summer of AI”.
This includes features such as “Sales Email Generation”, “Reply Recommendations” in Service Cloud, and “Subject Line Generation” in Marketing Cloud.
While Salesforce already has a huge amount of productivity features, and professionals around the globe continue to work hard to support user adoption, sometimes it’s hard to sell the value to those who don’t like to update a CRM.
But as AI on the Salesforce platform evolves, and new features allow manual work to be slowly eliminated from users’ daily lives, general happiness and productivity should increase.
According to a recent survey from Salesforce, business leaders see this having the following effects:
- 82% of business leaders say generative AI will lower overall business costs.
- 80% of business leaders say generative AI will increase revenue.
The compounding impact of AI and its potential to influence on a wide scale is pretty incredible. More productive users, lowering costs, increases in revenue… This could all usher in a new age of growth around the globe.
5. The Future of AI Will Be Fairly Mundane
Now, don’t get me wrong, Generative AI is very exciting; it’s a new technology which most of us have never experienced before. But so was the internet once upon a time – remember the first time you had a video call with someone? Or the first time you watched something on YouTube? All very thrilling moments! Ultimately, Generative AI will become just another tool we use to complete our daily tasks.
Although it’s a nice thought, I don’t believe we are entering the new age of utopia, whereby AI can do our jobs for us and allow us to work just a single day a week. Generative AI will most likely slot into our library of productivity apps in much the same way as email, CRM, Slack, and Zoom.
I make this point not to downplay the advancements in technology that are happening, but to hopefully calm the polarizing views of AI. Yes, we will all need to get behind Generative AI, learn prompt engineering, and ensure we can support our users with these new technologies, but training resources are already widely available (and coming soon to Trailhead!).
Whilst it’s impossible to predict anything in this fast-moving world we’re living in, there is no doubt that AI is here to stay, and it will embed itself in both our working and our personal lives.
Salesforce professionals are in a unique position; we will be using AI ourselves (including implementation features and tools), as well as supporting our users. But as it’s still early days, the real power of AI has yet to show itself. So, whilst we all need to get ready for this wave of transformation that is about to hit us, we have time to learn Generative AI in the right way, and ensure we are using it correctly.