Scared tech workers are scrambling to reinvent themselves as AI experts | Hacker News Reader

Scared tech workers are scrambling to reinvent themselves as AI experts

vox.com

39 pointsdacohenii23 comments

Open articleView on HN

It’s not just tech workers, it’s also tech projects.

Compare how minio has advertised itself through the years:

2023: High Performance Object Storage for AI [0]

2022: High Performance, Kubernetes Native Object Storage [1]

2020: MinIO is a high performance object storage server compatible with Amazon S3 APIs [2]

I get that buzzwords and marketing are important, but it rubs me the wrong way

[0] https://web.archive.org/web/20230628165639/https://github.co... [1] https://web.archive.org/web/20220204012915/https://github.co... [2] https://web.archive.org/web/20200528195457/https://github.co...

literally who?

“S3 clone” is easier to remember though.

Uhh, they were using AI moniker in 2019. They just circled back to it this year:

https://web.archive.org/web/20190428005007/https://min.io/

If you can rebrand yourself as an "expert" in some shiny new thing by spending a few weeks at a boot camp, there probably isn't much there and you will be competing for the same jobs against countless others who did the same.

I managed to completely ignore the cool new thing for the last 20 years and instead slowly built up knowledge in those problem domains that actually interest me, and it's worked out alright.

You assume that "branding yourself as an expert" is the same as "becoming an expert".

My read of his comment is the opposite of that.

Agreed. Of the many things I've learned over the years, two related ones are:

if the work you're doing is work that you're genuinely interested in, you have a huge immediate advantage, and

Running into where crowd is running out, and running out of where the crowd is running in usually (but not always) gives you a large advantage.

Yeah what most companies are even thinking about right now is prompt engineering which doesn’t require an advanced degree.

Actually developing models is a much smaller field (in terms of jobs) and the overlap with software engineering is small. I’m sure you can make a lot of money in it, but you’re better off coming at it from data science.

A UK perspective on this is that the AI trend, as described in the article, would quickly turn into a class divide. Many employers in AI/ML spaces seem to want postgraduate degrees. Often the academic entry requirements to such degrees aren't that high (usually a bachelors degree with an "ok" mark like a 2:2 or a 2:1 min in a related subject - not exactly sky high attainment criteria) but the financial barriers are relatively severe - far poorer finance options are available as compared to undergraduate degrees especially.

The dynamic then seems to be that while software development has been functioning as a relatively effective source of social mobility, if you can do the work you can create a decent career, the current AI wave threatens to pull up the ladder for many if this trend carries on as described.

That said I'm not too worried about it playing out like that yet, at least not in the medium term. I think industry is at some pinnacle of optimism around AI but don't quite realise that it's not a great fit for a lot of commercially relevant use cases like maintenance of existing code bases of more than trivial complexity and it's unlikely to be for a good while and where there is sensitivity to copyright. I think once that reality sets in there will be some degree of back-peddling.

I don't see it quite like that. It may be true that many employers are looking for postgrads. I don't know. But since practically _every single company_ large or small wants to know how or if they can take advantage of AI right now, there are plenty of opportunities to go around for anyone who has any knowledge or skills in generative AI.

And by far the most deployed technologies, even for postdocs, are going to be basically off-the-shelf or somewhat fine-tuned existing models via API or tooling being used for text and/or image generation for specific applications. In other words, things that in no way require the degree. That isn't to say that having real machine learning specialists isn't desirable, but there is plenty of room for people who just know how to apply the tools rather than invent new ones.

Actually there are plenty of "academic" papers coming out from people with advanced degrees that are mainly just applying GPT-4 to write code in some DSL to solve some category of problem. That stuff doesn't even require basic understanding of neural network architectures.

The latest models are so powerful it seems to be having a democratizing effect rather than increasing the class divide. From my perspective.

> the financial barriers are relatively severe

Anyone taking out loans for a PhD in the UK is a fool. Not only are our PhDs generally less well regarded than those from the EU and US, but there is also an abundance of grant funding available. If you can't get paid as part of a grant (and hence have fees taken care of) for a PhD here, then either the field of study has little economic value, or you are not considered good enough.

The reason why AI/ML spaces want postgraduate degrees is because there is a massive glut of candidates relative to openings. One of the things we realized over the past decade is that if you are not in the business of hiring AI researchers even companies with >1,000 SWEs may need only a dozen ML experts.

Even at fairly mature companies, fully utilizing a top tier ML expert may require the equivalent of 6-10 SWEs.

Anyway, while no one knows what the future holds, right now it doesn't seem like many companies are looking to actually take on AI as a core capability, they're looking to leverage open source tools. I'm at a large company right now that's betting pretty heavily on AI and we're thinking of bringing on 1 maybe 2 ML specialists over the next year. My friends at other large tech firms report similar plans.

Good lord, AI engineers get paid a median salary of $243,500, while non-AI engineers only get paid a median of $166,750, how can it be so much more? Is it because AI jobs are more concentrated in big tech?

There are way more non-AI engineers than actual AI-engineers.

The knowledge is harder to come by, and the field is (relatively) new. You need a fusion of data engineering, math, software engineering, and data analysis. Plus the ability to read and implement esoteric papers hot off the press, or in some cases innovate a new technique/module/architecture that hasn't been done before.

Oh, and you need to be able to do all the above _quickly_.

What's nice about it versus SWE is that you can totally ship hot garbage, doesn't have to be optimal at all, just "fast enough" and cheap enough.

half of them likely were crypto experts year ago.

Just wait for the blockchain AI “thought leaders” to come out of the woodwork.

There was already a pitch for distributed blockchain AI training before the crypto craze.

The deeper you dive into crypto code and structure, the more you realize it doesn't really work outside of some niches.

Modern ML has been totally different for me. It was janky before the craze, it is still janky as heck, and the ML community has its own unique blend of manic, reckless myopia and scammers. But the deeper I go, the more mind boggling it gets.

Or js “experts” cloning repos and claiming glory.

My problem with AI is that I just don't think it does the advertised task well. It really annoys me that the tech space in my area is increasingly 'embracing' AI* instead of things like Nix, Rust, Haskell, which are doing really cool things in the broader tech sphere.

*On that note: the phrase 'embracing AI' sounds so slimy. Why do I need to hug the robot?

I'd be interested to hear about the experiences of someone who has gone through that Berkeley program or a similar one.

Reply on news.ycombinator.com