A few tech predictions for 2023

I think humans overvalue their abilities to see the future and I’m hardly an exception, but allow me to partake in the fun. I will keep this up and you will be able to roast me throughout the year if I’m wrong.

It will be the dawn of a new era of AI-based products

OpenAI is leading the creation of a set of cloud-based primitives that are easy to understand even by non-technical people, who will dream of new products and applications for their domains. Likewise, the bar for developers to integrate these tools is constantly being lowered - although at a cost. Much like Amazon Web Services or Stripe are, paraphrasing Ben Thompson, like taxes for doing business on the internet, so will OpenAI and other providers take a (literal) toll on new AI-based businesses.

The paradigm was already shifting from “AI science” to “AI engineering” (many “AI startups” of the last few years were using ready-made models and technologies, and simply tuning them and applying them to their use case), but 2023 will see the trend grow exponentially thanks to the further “productization” of models.

The citizen developer movement will not explode (yet)

The vision holds all its initial potential but requires a multi-pronged and universal push involving corporate politics, compliance, security, training, etc. that a single entity will not be able to solve. We will see interesting initiatives or companies built around this in 2023, and we will likely see linear growth, but a year won’t be enough time to fulfill the initial promise of seeing less technical people make an impact in the development of enterprise solutions, at least at a massive scale.

Low-code will grow, but devs will be the main drivers

As a counter-point to the previous point: low-code tooling will see a growth spurt, not driven by “citizen developers” but by increased developer adoption of technologies that offer a good developer experience and capabilities to integrate well into “traditional” development workflows and existing software architectures. Tech like Retool, Plasmic, BuilderIo, PowerPlatform, HoneyCode, new-gen ETLs, various cloud offerings, etc.

Funding for startups will keep drying up, even if inflation is curbed in the coming months

Rising interest rates have been pushing investment mass away from startups and into the other end of the spectrum: good old bonds and other fixed-rate assets. Inflation numbers are looking better in the US as of this writing but even if interest rates do not rise, investment allocation is like a supermassive pendulum, and we will be seeing the 2022 fears materialize throughout most of 2023.

The great cleanse of zombie companies might not happen

Following the above, many are predicting that “bullshit companies” will fail en masse, but so much money has been pumped into startups in the last ten years that, I dunno, man - we may continue to have zombies (eg, 25 SaaS products for the same use case, thousands of crypto projects of dubious value, endless consumer tech and media products vying for people’s time, etc. ) for a long while. Especially if the Fed halts interest rate rises and we end up being in a not-here-nor-there scenario in regard to inflation vs. interest rate tensions.

ChatGPT and AI-assisted coding will not replace programmers (yet)

The bitter non-tech people who are expressing uninformed opinions on its code-generation capabilities, observing from the sidelines and waiting for some sort of comeuppance, are not going to get it just yet. Engineering and programming are much more than just coding. ChatGPT and tools like GitHub Copilot will allow programmers to be more productive and deal with less BS, but if programmer employment prospects worsen in 2023, it will be because of economic rather than technological factors. Software keeps getting more complex and continues to eat the world; AI coding technologies will be playing catch up, at least for a while - until we reach Singularity, or what-have-you.

There will be a rising movement of AI-assisted software dev methodologies

Following the above, there will be an increasing body of theory and training on how to use these tools efficiently - and the UX/DX of the tools will improve accordingly too. I love Copilot, but sometimes I’m finding that I get into messes that are easier sorted either by hand or by re-engineering prompts from scratch.

Microsoft will continue asserting its strength in the enterprise space

Their tech ecosystem is hyper-expansive, touching everything related to software development and IT (C#, .Net, TypeScript, Github, VSCode, PowerPlatform, Azure, the influence on OpenAI, etc.). Some of these projects and products are either already standards or are growing significantly. One could in theory build complex software projects without leaving their stack, from code editor to servers - not that I think this is necessarily a good idea.

JavaScript will continue to be the wild west

2023 will mark nearly three decades since JavaScript was made a front-end web standard, and fifteen years since it was made widely available as a back-end option through Node.js. I love the language and a good part of its open-source ecosystem, but it’s baffling to me that after all this time, we don’t have a standardized, batteries-included way of building full-stack JavaScript / TypeScript apps like Laravel in PHP or Django in Python. Some projects backed by insanely talented people are trying - just like others tried in the 2010’s - but as it stands there is no real leader bringing a “Rails-like” experience to JavaScript full-stack development. I am optimistic and I think that we will see progress in this area, but a year won’t be enough time to see definite progress given the current state of things.