A question in the Slack channel was stumping field CTO Nick Durkin’s team at the software deliverer Harness: how to tell Microsoft’s application development tool the location of certain .zip files.
To get a bit technical, the Microsoft Build Engine couldn’t find NuGet packages. One employee had a suggestion:
Have you asked ChatGPT? It’s been amazing with nearly all of my code- and infra-related questions.
ChatGPT, the generative-AI bot launched by OpenAI, did, in fact, answer the question—ultimately providing a “decent answer,” according to the asker, one not easily found in available documentation. The automated reply also sent example code-snippets.
“You can tell MSBuild where your NuGet packages are stored by specifying the package source in the NuGet.config file or by specifying the ‘-source’ option when running the ‘msbuild’ command…” read the response.
“It handled it in all of 30 seconds,” said Durkin.
ChatGPT has offered another valuable tool for coders, who are increasingly incorporating the AI-based language-model and all-around super-Jeeves into their everyday practices. The increased access to answers from artificial intelligence also potentially levels the programming field.
“This can make good developers great. This can make great developers excellent,” Durkin told IT Brew.
OpenAI season. This month, Microsoft announced that Azure enterprise customers will have access to the OpenAI-developed artificial-intelligence model, for applications ranging from summarizing customer data to writing reports or code. The general availability of Azure OpenAI grants customers access to OpenAI’s AI models, including GPT-3.5, Codex, and DALL-E-2.
While ChatGPT, the sometimes controversial chatbot trained on vast amounts of data from the internet, also has the ability to answer hackers’ more malicious questions, wider access to the model may help coders complete routine and time-consuming tasks, like finding errors.
Top insights for IT pros
From cybersecurity and big data to cloud computing, IT Brew covers the latest trends shaping business tech in our 4x weekly newsletter, virtual events with industry experts, and digital guides.
“It can help us in the conversion of code, potentially. It could help us testing code, because when you find an issue, AI could go in and work out where the issue might be,” said Tim Jones, managing director at the business-software company Advanced.
Artificial intelligence has increasingly supported application-development efforts. Copilot, a cloud-based AI tool developed by GitHub and OpenAI, offers code suggestions in real time.
Cameron Turner, VP of data science at the consultancy Kin + Carta (a Microsoft Gold Certified partner) sees the increasing availability of ChatGPT as a “flattening enabler” as the language-model AI becomes widely available in a “pay-as-you-go” model.
“You don’t have to have an army of AI engineers to go build these things and an army of mathematicians to tune them,” said Turner.
“If I’m starting a bagel shop, I can do just as well by my customers through those technologies. I can have the same level of accuracy and automated speech recognition, for example, in my drive-thru that Siri has, or that Alexa has, or Hey Google,” Turner added.
As requests come through the window, in other words, AI lets companies take orders quickly and, at the very least, provide a decent answer.—BH
Do you work in IT or have information about your IT department you want to share? Email [email protected].