Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The Rise of the AI Engineer

latent.space

Post reply on HN
Print view View original

41–50 of 159 posts

Re: The Rise of the AI Engineer

#41

Post by sanderjd »

Der_Einzige wrote: post #8

If you are someone who was working in LLMs/Generative AI before it got "cool", The market has fundamentally changed and changed extremely in your favor. Talent is extremely scarce and expensive right now in this space. Ask for well above FAANG compensation. Try to shop around for offers. I had 4 offers and a bidding war for my talent (with no leetcode - but I have an extensive github and publications about LLMs), whe…

I'm someone who would like to get "really good really fast", but have found that the on-ramps remain pretty weak. For instance, I have yet to find a good book on the subject. There are a ton of tutorials and articles, but it's maddening to try to cobble together any depth of understanding from those little nuggets. And there are tons of good papers on how the systems actually work, but these are not very useful for people who are new to this set of tools to figure out how to use them to create useful things.

But I think this will hole will close up incredibly quickly over the next six months. (And it's yet another really good opportunity for people like you to be involved in that gold rush!)

Edit to add: For instance, the article lightly lambasts peoples' recommendations to read up on AI / ML fundamentals, and contains this:

> n the near future, nobody will recommend starting in AI Engineering by reading Attention is All You Need, just like you do not start driving by reading the schematics for the Ford Model T.

But, amazingly, it doesn't actually suggest an alternative starting point*. There is this call to action about the conference, but presumably that will mostly be people who have already figured this out to some degree. But (to carry on the author's analogy) what is the recommendation for driver's ed?

Re: The Rise of the AI Engineer

#42

Post by minimaxir »

I would never be able to be hired as an "AI Engineer" or "Prompt Engineer" despite my extensive AI portfolio because my job title is Data Scientist and the discrepancy would confuse most hiring managers.

Job titles are moving faster than career ladders.

Re: The Rise of the AI Engineer

#43

Post by wpietri »

minimaxir wrote: post #29
voz_ wrote: post #9

Earlier quoted context omitted.

Imagine comparing blockchain and AI.

Unfortunately, web3 bros pivoted to AI, which just adds even more noise to the space.

For me that's absolutely the strongest ground for comparison.

In recent decades, we've had two big waves of tech advance: the Web and mobile. A lot of people have lived through them both, giving them an expectation that another such wave should be along soon. You could see that in the decade of blockchain/ICO/DAO/NFT/web3 hype, where people, many with shaky credentials, touted the transformation soon to come, taking in a lot of cash.

In retrospect, from Mt Gox to FTX we can see that it was all horseshit. The main real advance was decentralizing not finance or property or computing, but the Ponzi scheme.

Despite this failure, and echoing The Great Disappointment [1], we see a lot of the same hype and even the same people. Is it possible that this is different, that there's more substance here? Or is this going to be another Groupon or Metaverse? That does not matter to the hypesters. They're going to run the same routine that worked before. They're going to take in a lot of money, which is the primary goal. It's possible that some of them will, by luck or accident, latch on to something that isn't a total fraud. Surely most of them won't.

But we should never forget that is basically irrelevant to a lot of people in the early stages of a cycle. And not just for the fraudsters, but for anybody who makes their living on the upswing of the the hype cycle, including a notable fraction of investors, "experts", and journalists. how ever it turns out, they'll get paid just fine.

[1] https://en.wikipedia.org/wiki/Great_Disappointment

Re: The Rise of the AI Engineer

#44

Post by minimaxir »

wpietri wrote: post #43
minimaxir wrote: post #29

Earlier quoted context omitted.

Unfortunately, web3 bros pivoted to AI, which just adds even more noise to the space.

For me that's absolutely the strongest ground for comparison. In recent decades, we've had two big waves of tech advance: the Web and mobile. A lot of people have lived through them both, giving them an expectation that another such wave should be along soon. You could see that in the decade of blockchain/ICO/DAO/NFT/web3 hype, where people, many with shaky credentials, touted the transformation soon to come, taking…

Web3 / Metaverse was a solution in search of a problem.

AI is solving problems today.

Re: The Rise of the AI Engineer

#45

Post by swyx »

cj wrote: post #14

I can see a world where "ML Engineer" (or similar) is someone that's hired to solve a known problem (whether it be with classifiers, LLMs, neural nets, etc), whereas a "AI Engineer" (or whatever the title) is hired to figure out how the hell to capitalize on the AI hype, without a specific problem to solve. IMO right now we're entering the "Peak of Inflated Expectations" in Gartner's hype cycle model. https://en.wiki…

(author here) There are at least 4 killer apps ($100m/yr revenue potential) so far:

1. Generative Text for writing - Jasper AI going 0 to $75m ARR in 2 years

2. Generative Art for non-artists - Midjourney has by some accounts $80m ARR

3. Copilot for knowledge workers - GitHub’s Copilot has roughly 50-80m ARR as well

4. Conversational AI UX - ChatGPT probably has >$100m ARR by now, Bing Chat has brought $m's worth of attention to Bing

(More on agents as the immature #5: https://www.latent.space/p/agents)

what else do you need to see to believe? (geniune question)

Re: The Rise of the AI Engineer

#46

Post by janalsncm »

sanderjd wrote: post #27
voz_ wrote: post #9

Earlier quoted context omitted.

Imagine comparing blockchain and AI.

I would submit that we don't yet have enough evidence to say whether the comparison is apt or inapt. I personally think what's going on with these "generative" models seems like a bigger deal than blockchain, but it's fiendishly difficult to know what is or isn't hype while embedded within a hype cycle.

Seriously? The only comparison you can make is that both were hyped. Digging even a millimeter under the surface reveals they’re completely different.

Blockchain was and still is rife with scams and “you just don’t understand the technology bro” hype men. Just check out Dirty Bubble Media. Blockchain was rarely if ever a product that solved a problem, the whole point is The Line Goes Up. That’s why no one uses blockchain in industry, and crypto bros were constantly finding themselves proposing silly use cases like ticket sales and property deeds. These are people who have apparently never heard of a relational database.

The hype around AI is due to increased attention on things that have already existed and have already been studied, used, and improved for decades now. There was never much R&D into blockchain tech because the tech isn’t the point. For ML, there are researchers who have worked on these problems for decades. It doesn’t need to justify its own existence, the justification is that it can solve real problems.

Re: The Rise of the AI Engineer

#47

Post by wpietri »

minimaxir wrote: post #44
wpietri wrote: post #43

Earlier quoted context omitted.

For me that's absolutely the strongest ground for comparison. In recent decades, we've had two big waves of tech advance: the Web and mobile. A lot of people have lived through them both, giving them an expectation that another such wave should be along soon. You could see that in the decade of blockchain/ICO/DAO/NFT/web3 hype, where people, many with shaky credentials, touted the transformation soon to come, taking…

Web3 / Metaverse was a solution in search of a problem. AI is solving problems today.

You're responding as if I said they were the same, but I was pretty careful to say the opposite. My point is explicitly setting the utility of "AI" aside.

Re: The Rise of the AI Engineer

#48

Post by echelon »

swyx wrote: post #45
cj wrote: post #14

I can see a world where "ML Engineer" (or similar) is someone that's hired to solve a known problem (whether it be with classifiers, LLMs, neural nets, etc), whereas a "AI Engineer" (or whatever the title) is hired to figure out how the hell to capitalize on the AI hype, without a specific problem to solve. IMO right now we're entering the "Peak of Inflated Expectations" in Gartner's hype cycle model. https://en.wiki…

(author here) There are at least 4 killer apps ($100m/yr revenue potential) so far: 1. Generative Text for writing - Jasper AI going 0 to $75m ARR in 2 years 2. Generative Art for non-artists - Midjourney has by some accounts $80m ARR 3. Copilot for knowledge workers - GitHub’s Copilot has roughly 50-80m ARR as well 4. Conversational AI UX - ChatGPT probably has >$100m ARR by now, Bing Chat has brought $m's worth of…

The longer HN continues to doubt in AI's deliverables, the bigger our revenues and our moats can grow. Don't tell them.

(It's not that hard to hit $1M ARR with a good AI product. So many classes of new products and solutions have opened up.)

Re: The Rise of the AI Engineer

#50

Post by swyx »

vsareto wrote: post #5

For this conference, I think it would also be great to have a presentation about job requirements at various levels, then continue to cite that presentation often. It's important to have an on-ramp for people interested in this space with well-known requirements to aim for. This also makes it easy for companies to know what they're getting with this title. The ambiguity of full stack engineers meant that it was a dum…

(author here) one of my ideas of research for this blogpost was to actually ask a few employer friends who want to hire AI Engineers for a job description, and then put out a generic JD that people can use and clone (kind of a SAFE for AI Engineers if you think about it). however I feel like it is too early still to prescribe something. I'll maybe put up a strawman for the conference and would love feedback then.
Post reply on HN