Five-time tech business Founder and Chair of the Tenzing Entrepreneurs Panel, Bobby Healy is pretty excited about ChatGPT and its game-changing potential for tech businesses.
It’s 2023 and ChatGPT has just made headlines the world over. What most people don’t know is ChatGPT has been around since 2018. And some tech firms – mine included – have already encouraged their programming teams to use it with incredible productivity results.
As with the introduction of all new technology, businesses that adapt to ChatGPT first will have the advantage. I believe tech firms only have around two years to get ahead of the competition by making the most of this software’s powerful potential.
With six months ChatGPT experience under my and my company’s belt, I’m sharing my ChatGPT insights. Get ready to learn what ChatGPT is, how it works, how it’s changing the way programming teams operate and the transformative benefits it could bring to your business. Before finishing up with my top tips to help your business implement this impressive tool.
Read this first – what is ChatGPT?
Ask ChatGPT what it is and how it works, and this is what it says: “ChatGPT, or Chatbot Generative Pre-trained Transformer, is a sophisticated AI model developed by OpenAI based on the GPT-4 architecture. It’s designed to understand and generate human-like text, enabling it to have conversations with users, answer questions and provide relevant information.”
Developed through an iterative process which involves pushing huge amounts of data through a large language model, ChatGPT is a highly complex algorithm that works out the right results from random information. Similar to the way humans learn, its outputs are weighted and adjusted in response to correct results.
ChatGPT layers a web-based, chat user interface over the top of its machine learning model, making it intuitive to work with. By using general-purpose transformers – the GPT in ChatGPT – the platform is a strong all-rounder. It can learn languages or any other type of information within its training data set.
In the same way human brains produce answers based on the information they’ve digested throughout their lives, ChatGPT produces answers based on the data it has been trained on.
Unlike humans, ChatGPT’s experience is the sum of everyone’s experience – it’s the experience of the internet. Giving it a more comprehensive understanding of the world than any individual.
ChatGPT – the need-to-knows
- First launched in 2018 as ChatGPT-1.
- Current version is ChatGPT-4.
- You can enter a maximum of 25 inputs in 3 hours in GPT-4 although this is a temporary restriction.
- It’s been trained on trillions of data points up to 2021 so it doesn’t have any knowledge beyond this date but you can use plugins to address that.
- ChatGPT is more useful for new programming stacks than old. Stacks are the technologies used to develop a piece of software – because old stack coding isn’t as prevalent on the internet developer websites such as Stack Exchange, ChatGPT hasn’t learned it yet.
- Businesses can use plug-ins that allow ChatGPT to access organisational data. It can then respond to business-specific inputs and solve organisational problems.
What does this mean for tech firm leaders and programmers?
For the past few years, programmers like me have been experimenting with ChatGPT and we’ve discovered ChatGPT and coding go hand-in-hand.
Because code is based purely on technical design, rules and logic, ChatGPT is able to follow the syntax and reasoning of the programming languages it’s trained on. It rapidly generates code that’s highly likely to be error-free. Which has huge significance for the tech world.
From my experience working with ChatGPT in my business, I can see four key areas it will impact – people, process, product and profit – with the following benefits.
Programming quality, speed and cost will improve immediately
When a programmer writes a piece of code from scratch, even the most skilled individuals might not achieve perfection the first time.
Enter ChatGPT. Capable of writing flawless blocks of code which will be right most of the time, ChatGPT dramatically accelerates the coding process.
Code that might take me a day to write can be written by ChatGPT in 30 seconds. Making software development much quicker. Platforms will also be more reliable due to the cleanliness of the code, reducing project timelines and shrinking the total cost of ownership.
Programmers will become fully productive across multiple stacks overnight
Good programmers who are also multi stack developers are highly valuable because they’re rare.
But with ChatGPT, a programmer who’s really good in one programming language can immediately excel in other languages and environments. Suddenly, your C++ programmer is now a good Java programmer too.
This will enable experienced programmers to quickly migrate across different platforms and become much more flexible in terms of the work they can carry out. The granularity of your team is much finer. This is particularly true for smaller teams.
The make-up of your programming team will change
Your programmers will be able to outsource most new atomic code (standalone work that doesn’t need input from anyone else) to an infinite number of high-quality ChatGPT sessions.
This will free your teams to focus on building blocks of code that work logically together. As a result, senior programmers will become way more valuable. Less experienced programmers less so.
You’ll still need System Architects to design scalable systems. But all the plumbing, configuration and scripting will become much easier or disappear. And SQL gurus – who use the decades-old Structured Query programming language to interrogate databases – will become a thing of the past. Just give ChatGPT a connection to your database and it will do what your SQL guru does, only much faster with fewer errors. And I say this as an SQL guru myself!
With more productive system admin and DevOps teams, you won’t need as many people in these roles. Alternatively, you could choose to build, launch and manage more products. Or something in between.
Upskilling people for senior programming roles
Lead programmers’ current roles won’t change that much. They’ll still collaborate with the business, product development and everyone else they talk to today. But they won’t need to look after some or all of the people who report into them now because these humans will be replaced by ChatGPT.
You’ll still need a handful of less experienced colleagues and you’ll accelerate their development using ChatGPT. The timeline on this has shrunk massively. But only those who are truly passionate about programming will be able to make the leap to the demands of the senior role – reengineering existing code or designs.
Programmers who like to sit and code all day have a short shelf life. It’s likely they’ll end up working for themselves where they’ll make a good income, using ChatGPT to rapidly code multiple projects concurrently.
You’ll get minimum viable products to market more quickly
Startups without a technical founder typically need to hire a number of programmers to build their idea into a minimum viable product (MVP). This can be expensive. But money isn’t really the issue for many firms – speed is.
It’s virtually impossible to find talented engineers willing to drop everything for three months to build your proof of concept. Get ChatGPT on it and you can start pushing stuff through right away, from the MVP to the company website, marketing collateral and social media posts.
This speed-to-market also applies to more mature firms with new products to build and stretched resources. Introduce ChatGPT to your teams and watch them rapidly build high-quality products ready for the measure and learn stages.
ChatGPT will become your latest marketing opportunity
Companies that produce or promote application programming interfaces (APIs) need to make sure their information is in ChatGPT’s training data. This will make it easy for developers to adopt your API because ChatGPT will produce perfect code for them to use. With your code in an organisation’s platform, it’s much harder for them to step away.
This is an exciting marketing opportunity. But how do you get ChatGPT to spit out example code using your API and not your competitor’s?
Think of ChatGPT as similar to the early stages of Google and search engine optimisation (SEO). The more references Google can find on the internet to your product or service, the higher up the Google pages your website will appear. ChatGTP is similar.
Let’s pretend you run an online payment business. When someone asks ChatGPT to write some payment code, you want ChatGPT to give an example using your code. The more of your data you can get into its machine learning database, the more likely ChatGPT is to reference it.
In the early days of my first business, we wrote hundreds of web pages robotically to convince Google into putting us at the top of search results. I’m predicting a similar volume race as businesses seek to influence ChatGPT’s results.
Visual representations of data can be produced quickly and easily
It’s not only your programming teams who can leverage ChatGPT. Pulling together reporting dashboards or other visual data representations used to take a lot of effort and specialist knowledge. But did you know, ChatGPT can build analytics dashboards and more?
Microsoft is already embedding ChatGPT into all their productivity tools. So if you’re using Microsoft’s analysis tools, you no longer need to be an expert. You can just say: “Here’s my question.” And ChatGPT will deliver the answer for you.
From geographic information, video and image editing to animation and analytics dashboards, ChatGPT can analyse huge swathes of data and produce all the sexy interfaces you want. Helping everyone in your business produce more complex outputs and extract more value from their tools. Which will reduce the time and cost spent producing and maintaining visual data and minimising your reliance on experts.
Essential factors to consider for your business
As with any emerging technology, businesses that proactively embrace it will gain a competitive advantage. In companies where leaders and managers are willing to bet on ChatGPT, I see the transformation I’ve described here happening in a single generation of projects.
Your current projects and organisational structure won’t change. But, when the whole organisation embraces ChatGPT and spins up new projects, tasks that might previously have taken a whole team a year or two to complete will take one person one month.
This transition will happen in the next two years, perhaps even sooner in companies where top-level execs are driving immediate ChatGPT adoption. For organisations not led by programmers, there will be a bit of concern and the transition will take longer.
These are my tips to help you leverage ChatGPT.
Get your people using ChatGPT today
We’re at the start of a new tech arms race with ChatGPT. So you need to get your people – in programming and beyond – up to speed as quickly as possible.
There’s no need for formal training. Just get in there and start playing with the software. As a programmer, that’s the way I learned to use ChatGPT. And it’s helped me deliver work way more quickly and accurately than I could have on my own.
The key to taking this forward is sharing your teams’ learning. In my company, we have a dedicated Slack channel where everyone shares useful or interesting ChatGPT examples. Including things you might not have thought about before.
For example, one of the team shared a bit of code ChatGPT had created which turns an Excel spreadsheet into a java enum. Programmers need to use these ‘enums’ – or lists of information – for their work. But creating them manually involves a tedious bit of editing which is prone to error. I hadn’t thought about using GPT in this way before. So it’s a continuous education process where you’re learning new tricks all the time.
Be prepared for resistance
As a programmer, I know we can be a stubborn bunch. So, you’ll find some people resistant to change. The reality is, they’ll quickly end up being left behind. Until they see how everyone else is using the platform – then they’ll rapidly catch up.
Think about customising ChatGPT to your business
ChatGPT is already pretty incredible. But you can make it even more useful by using plugins so it can read your company’s data and train on your business.
I envisage a company that already has your data – like Microsoft – will allow you to turn on a private ChatGPT room for your company and team. Training ChatGPT on your data so it can solve problems within your organisation’s specific context.
I strongly believe the dawn of ChatGPT marks a new era for programming teams and tech companies. But only if CFOs and CEOs harness the power of this innovative technology. Choose to seize the moment and embrace ChatGPT and you’ll revolutionise your operations, enhance productivity and unlock exciting growth opportunities. Ensuring you establish a strong competitive edge and play a role in reshaping our industry’s landscape.