> Brynjolfsson analyzed millions of ADP payroll records and found a 13% relative decline in employment for early-career workers (ages 22-25) in AI-exposed occupations since late 2022.
> So what’s the mechanism at play? AI replaces codified knowledge
Many job postings peaked in 2022 due to the pandemic. The original paper tries to account for this but falls short in my opinion.
Original paper said[1]:
> One possibility is that our results are explained by a general slowdown in technology hiring from 2022 to 2023 as firms recovered from the COVID-19 Pandemic...
> Figure A12 shows employment changes by age and exposure quintile after excluding computer occupations...
> Figure A13 shows results when excluding firms in information technology or computer systems design...
> ... These results indicate that our findings are not specific to technology roles.
Excluding computer and IT jobs is not enough in my opinion. Look at all these other occupations which had peak hiring in 2022.
It's not free if you have to trade your info for them. It's not like I have a business case for photos of broken insulators, just trying to check what you made.
Did claudebot have paying customers? My understanding with these companies is that you buy the market, the product can just be forked (like Amazon did)
There's already Zulip, Mattermost, and many others. Building a chat application should be considered a tar pit problem IMO. A lot of success relies on network effects and familiarity, and the product looks deceptively simple.
It's unlikely you can build one that is better than Slack without years of investment. Even if you do, it's still an uphill battle.
> Building a chat application should be considered a tar pit problem IMO
Yes. For example Discord originated as a side-project for a team who were supposed to be building an MOBA. That’s why if you try to build a discord chatbot or custom command or whatever, the servers are called “guilds” etc.[1]
Slack was also developed by a team who were supposed to be developing a video game.[2]
I recently started looking into Zulip and while I can see that it is a complete product its mobile UI is so cluttered and funky I don’t understand how anyone could use it. The desktop web UI seems OK but try this on your phone: https://zulip.com/new/demo/
They have the iOS Safari problem with the keyboard and body scroll, tiny icons, super busy UI. I was hoping to help some folks move off Discord to something else and Zulip is not what I would volunteer to do support for when the users are not techies. Heck, as a techie my eyes glaze over looking at it. I really wish it was slicker and more usable but it simply isn’t.
Zulip is pretty weird compared to the rest, it's always hard to tell what's even going on with threads within threads within threads. Far more experimental than all others which are basically all the same.
There's also Discord of course, but they've recently announced their impending implosion.
Author here. You guys are reacting like engineers - it's not the raw features, it's the critical mass that only a rare few like openai can attract.
I don't care that someone else is already trying to build a slack killer. They do not have critical mass.
> A lot of success relies on network effects and familiarity, and the product looks deceptively simple. It's unlikely you can build one that is better than Slack
i agree that you and i can't build one. openai can. article argues that because it can, it should.
There's network effects and then there's core competencies. OpenAI has not demonstrated their ability to create software that is not a primary use case for LLMs. Chat is absolutely not a primary use case for LLMs, and so far LLMs have been sold as a value-add for traditional software.
The argument that OpenAI has the critical mass to dethrone Slack can be made for just about any other product with an 800-pound gorilla market leader. Windows, Office, Photoshop/Premier, Search, GMail, Figma, etc. Thus far, we have yet to see OpenAI build anything like these at scale, and there's no reason to assume their successes in the LLM space will translate.
I agree that they should build killer apps like these, because they are at extreme risk of being commoditized by smaller, better, faster genAI systems, but I don't think anything they do currently shows that they can.
"You guys are reacting like engineers" is a very wave-y dismissal of the many practical questions raised about why exactly OpenAI should expand into a product that's tangentially related (at best) to their core competency of AI.
The chain of logic in the article is explicitly spelled out as: Sam Altman said OpenAI will grow into new products -> Altman says to tell them what these products should be -> You say: Slack sucks so.... how about Slack?
I think most people, engineers or otherwise, reading the article have an understandable reaction of mostly bafflement as to why we are even talking about this, specifically, to begin with?
Mattermost team here. Agree we could have done a better job communicating. The change started in 2023 and we had made a lot of effort to work with the largest unsupported deployments early.
Our understanding is that the organizations most impacted were those using the unsupported Mattermost commercial version, not the open source version. The commercial version of Mattermost is offered in Docker, K8, etc.
If you look into the license of the Mattermost instance you ran, what is the "Enterprise Edition" (i.e. commercial version that upgrades into paid offering) or under MIT license (open source licensed offering, bundled with GitLab omnibus)?
I've been told to wait for a pay increase/promotion twice. And I got it both times (luckily). The time periods were only a few months or a year each time.
I think it's a judgement call but making such a long-out promise like 3 years in the tech industry is a huge red flag. Even at one year you should be skeptical and asking how/why as the author suggests.
Being told you have to wait until the next pay review cycle, is normal. It’s how a business with healthy and defined processes should operate.
But you should only be waiting at most a year. If you get told “wait 2+ years” then that’s usually a sign that they’ve already decided you’re not eligible (for whatever reasons they decide) but don’t want to be candid with you.
If you get told to wait for any duration beyond the next pay review cycle, then take that as a sign that you’re not going to progress under the current regime.
There are obviously going to be exceptions. Every rules has that. Hear why I said “usually a sign” rather than “it’s a guarantee with out any exceptions”.
But to take your startup example, they generally short on base salary with the hope that you score big when the company sells / floats. Which is a very different scenario to saying “we aren’t going to pay you more because we are unprofitable”.
Also, if a regular (ie non-startup) business isn’t profitable and are then freezing wages as a result, then that’s another good indicator to update your CV. You might be lucky to get a decent severance package, but even if you do, you’ll still want that CV updated.
Because it shows the company has processes in place for economic planning, with budget allocation, and all the other systems and checks that are meant to ensure stability and profitability.
That’s not to say that businesses with these processes defined can’t still be total shitshows. But the ones that don’t have those processes are more likely to be shitshows.
There's such a long history of Heroku in this post, which isn't bad per-say, but the post is lacking any evidence or insight to support the title. I'm not sure if the point is to say "I know a lot about Heroku, so trust me" but it comes off like that.
> All I can say is: it sounds to me like there is hope, as a lot of these pains are being addressed actively.
If you're coming for the title, I think reading the above quote is sufficient.
> So what’s the mechanism at play? AI replaces codified knowledge
Many job postings peaked in 2022 due to the pandemic. The original paper tries to account for this but falls short in my opinion.
Original paper said[1]:
> One possibility is that our results are explained by a general slowdown in technology hiring from 2022 to 2023 as firms recovered from the COVID-19 Pandemic...
> Figure A12 shows employment changes by age and exposure quintile after excluding computer occupations...
> Figure A13 shows results when excluding firms in information technology or computer systems design...
> ... These results indicate that our findings are not specific to technology roles.
Excluding computer and IT jobs is not enough in my opinion. Look at all these other occupations which had peak hiring in 2022.
Nursing jobs in the US: https://fred.stlouisfed.org/series/IHLIDXUSTPNURS
Sales jobs in the US: https://fred.stlouisfed.org/series/IHLIDXUSTPSALE
Scientific research & development jobs in the US: https://fred.stlouisfed.org/series/IHLIDXUSTPSCREDE
Baking & finance jobs in the US: https://fred.stlouisfed.org/series/IHLIDXUSTPBAFI
[1] https://digitaleconomy.stanford.edu/app/uploads/2025/12/Cana...
reply