[gumboza]

- The slow death of SPAs and client side JavaScript (including WASM in that) will commence due to the dubious ROI.

- War against humans and AI commences. The AI content farms will destroy everything by reducing the signal to noise ratio so that all content is worthless regardless of who wrote it.

- Major cloud provider will screw something up and lose a lot of clients leading to a minor shift of corporates back to their own infrastructure.

[caldarons]

- War against humans and AI commences. The AI content farms will destroy everything by reducing the signal to noise ratio so that all content is worthless regardless of who wrote it.

Honestly, since the release of ChatGPT and Stable Diffusion this Us something I have been thinking about a lot. I we thought that what we saw in the last 10-15 years was an explosion in content available online, we cannot even imagine the quantity of content that will come. I am actually surprised this has not happened yet, although I think that we will see more and more articles with contributions by ChatGPT. Short to mid-term I hope that this will not lead to a decrease in content quality/lack of diversity. Long-term I think we will see mechanisms arise to distinguish human writing from AI generated content.

I am aware that this might sound pessimistic, I am actually excited to see where his things will evolve!

[CMCDragonkai]

Probably through digital signatures and sovereign identity.

[gumboza]

Adding more technology isn’t going to help this.

[brutusborn]

Why?

[gumboza]

You can sign garbage therefore it is a human issue.

[brutusborn]

But can’t you whitelist/blacklist signatures? Then use a web of trust to determine the probability of trusting non-listed signatures. Filter content by trust probability and spam should disappear.

[dmak]

> - The slow death of SPAs and client side JavaScript (including WASM in that) will commence due to the dubious ROI.

What do you think would replace it? it seems impossible to manage a modern frontend app without it. For most cases, I think we don't need it though.

[gumboza]

Server side rendering with light weight JavaScript. The point you make about most cases not needing it is spot on.

[Gigachad]

This is how most products start and then inevitably it starts to require more complex interactive pages so you have to add react and then it becomes easier for everything to be react.

Client side rendering is not strictly tied to single page apps. At a previous company we had rails render out a div containing all the data the page needed to mount the react component. No complex api design, no weird permission issues, etc.

[gumboza]

I’ve worked on extremely complex projects (many hundreds of endpoints, many TB of data). React definitely does not scale for those at all. It has a niche for a few use cases but it’s terrible for most I have found.

It’s all about applying the correct solution to the problem and that isn’t the web if it involves react.

[ojr]

React scales for Facebook level traffic, I doubt there are many use cases where projects need to scale React to serve more users than Facebook

[gumboza]

Facebook does a few very simple things lots of times a second.

It doesn’t scale to lots of different things regularly.

[dmak]

I've done that before, and it works well especially if that data doesn't frequently change or grow too big.

[egeozcan]

This sounds more like a wish than a prediction. I'm saying this as someone who used to love intercooler.js (and really dislikes htmx).

[gumboza]

It’s a necessity. The end user experience for React isn’t necessarily the best.

Where it exists, the web is probably the wrong solution for the problem.

[quickthrower2]

I predict the opposite: more movement towards code running in the browser. Not that it is necessary always a good thing!

[lgas]

I think we'll see both, with a sort of push to the two extremes initially, and then the two extremes coming back together to meet in the middle again as WASM gains popularity on both the client and the server leading to a lot more "isomorphic" (stupid term in this context, but whatever) codebases.

[dmak]

I highly doubt WASM will gain much ground. It may for games, but I am willing to take a bet that there are more web apps out there than games.

[mhuffman]

I think the eventual common use for WASM will eventually be DRM and related types of page code "security" and obfuscation. Basically a way for companies to "protect" their tech/IP from easy duplication while not having to carry the expensive burden of the cost of back-end processing in the cloud.

[egeozcan]

Is WASM harder to de-obfuscate than your typical transpiled/minimized JS soup?

[mhuffman]

From the perspective of salespeople and managers ... maybe!

[quickthrower2]

WASM allows you to run code “isomorphically” on the front and back end, which is a convenience that will be hard to resist.

[dmak]

That's not often a requirement for many companies. Most companies end up with a different backend technology for a diverse range of reasons.

[osrec]

I agree with this.

[SyzygistSix]

>War against humans and AI commences. The AI content farms will destroy everything by reducing the signal to noise ratio so that all content is worthless regardless of who wrote it.

That already began this year. We may realize it in 2023 though.