AI can make developers more productive

AI can make a developer more productive by providing planning and coding features.

These features are often integrated into existing editors, most notably as VS Code extensions.

If you’re an experienced developer, you’ll most likely be treating AI as a junior assistant who’s recently left college and has a wealth of knowledge but little experience. So you’ll want to monitor what it does closely and make sure it’s getting things right.

If you’re a less experienced developer, you’ll trust your assistant a little more, maybe more than you should, but you’ll rely more on it telling you what it’s doing and why. In this case

Either way, it pays to research a little while you work with the tools.

AI Tools

There are as many tools out there as there are days in the year, but here are a couple of my favourites:

Continue : Yes – it’s the name of the tool not an instruction to proceed. This is an open source VS Code extension that offers many of the features of the commercial products. If you need to work within strict company policies regarding security/licensing combine this with a locally hosted LLM running on Ollama and you should be good to go. Just bear in mind the hardware requirements and the fact that you probably wont be able to host anything as capable as some of the larger and more advanced models. But if them’s the rules, it might be your best option.

Cline : Another VS Code extension and one that offers more features. If you plan to use AI for personal use, this might become your best friend, assistant-wise.

If you have yet to dip your toes in the water, here are a few tips

• Get a feel for the tools with a personal project – you can use some decent tools at virtually zero cost without breaking any rules
• Find out about your company’s policy on AI tools. Do they have any licenses you can use to access AI that don’t break their security rules?
• If things are a bit restrictive (they wont let you use any paid tools and they wont let you talk to any external LLM (what drives the chatbots), you may need the local hosting option. Don’t feel sad at this point, it’s doable.

Let’s look at some tools that cost nothing, or net to nothing for personal use:

• Google AI Studio – In addiition to some useful tools, if you look at the top you’ll see an option to Get API key. That can be used in tools like Cline to get you some free access to powerful LLMs.
• Open Router AI This used to give you free access to some other good LLMs, but recently at time of writing, they will do that only if you have $10.00 credit (tip – I would put in $11 in case you accidentally use one of the paid models). I have found this to be very useful.
• Ollama – Just install it. It will give you the ability to run up LLMs locally with no cost (other than a bit of energy and RAM). Be aware that unless you have some sort of super computer or a very good GPU card, you’ll need to look for the smaller models, depending on your system RAM and processor.

Gemini Code Assist : One of the easiest ways to get started with using AI to help with personal coding tasks is the Google Code Assist extension in VS Code, which is really straightforward to use and quite effective. Simply find it in the extensions tab and install. Play around with it to get a feel for how code assistants can help improve your productivity. It recently increased its free plan (personal use) quotas to very generous levels.

Copilot in VS Code : This can be used in VS Code either for autocompletions, code assistance or in the recently added VS Code Agent Mode. It’s worth taking a look at this, because it offers similar functionality to Cline, but for personal use there is a free quota.

Have fun