Did AI Just Really Take Our Software Engineering Jobs? (Or Not?)

Key Takeaways

hey everyone today we're diving into the buzz around Deon the AI claiming to be the world's first software engineer been getting a lot of messages from people asking if they need to stop studying computer science or if software engineering is a dead end is AI going to replace all software engineers in the near future and people have been sending me interview of Jensen Wang who said software engineering is dead as promised I do have a lot of thoughts around this so let's see if that is really a revolution or is it just a clever marketing gimmick so I want to talk about five points that I want you to think about when you come across news articles like this so you can make better informed decisions about your future and maybe don't freak out as much so here we go why I'm not entirely sold on Deon number one is the limited benchmarks that they presented passing an interview or getting some contract work done on upw work doesn't always mean someone is really good at coding sometimes it just shows how well they can talk about coding and an interview deon's ability to pass interviews doesn't necessarily reflect how good Deon actually is at coding if anything it only proves how flawed or dumb the interview process is rather than how good Deon is at coding so I don't know if there is a way to really accurately measure how good Devon's coding skill is because of the the subjectivity of software engineering work you might think that software engineering work can be Quantified really easily but even the metrics that we use are pretty subjective and I talk extensively about this in my coaching program which focuses more on how to quantify your work and promote your work because of the subjective nature you can check out my courses if you're interested in talking more about that I think I'm going to give you a 10% discount code in the description so claim that 10% discount code back to my list number two is the nature of Contracting work often times the challenges with the Contracting work is not the work itself but is that the scope is not that clear here's a little story about me when I was 22 I left my full-time job at IBM and I tried living the nomatic lifestyle jumping from contract to contract for a couple of years and contracting jobs are quite different from regular full-time jobs so in Real World Companies you usually don't start a project from scratch almost never unless you're working for a preed startup that doesn't even have an MVP right so instead you often have to go figure out how to work with the existing codebase that's already there and one thing that I often have to remind new Engineers or Junior Engineers is that they always want to start projects from scratch because that's just so much easier but in the real world that's just not how it's going to work and you're more likely going to have to deal with old archaic dinosaur code and there are usually reasons why things are done the way they are in these older code there is a bit of a challenge and a learning curve to learning all of the ins and outs of existing code base which is not very straightforward my guess is that if we were to place Devon in a real company's code base like Facebook or meta I used to work at there's just not enough evidence to confidently support that Devon will succeed in a more complex environment like that okay number three is that writing code is the easiest part of being a software engineer yeah sure Devon can compare different options for building websites and apps which is useful but is it amazing or groundbreaking I don't really think so it seems that Devon is able to do some decent work or at least able to produce enough decent work for cognition to be able to select a few videos to highlight for marketing campaigns but in the eyes of anyone who knows anything about software engineering these videos are not very impressive I've been seeing even YouTubers saying that the future of engineering is going to change completely because of Devon all Engineers are going to become like managers Who oversee projects only here's the new spash guess what staff Engineers spend the bulk of their time working on architecture producing spec dogs mentoring other Engineers leadership discussions etc etc it's common for senior staff Engineers to not do much coding at all working on non-technical work has always been a big part of being a senior engineer so the more experience you have the less coding you typically do that's how it always has been so I don't really see how de is going to change anything for senior Engineers number four is the marketing gimmick that I mentioned earlier this honestly really goes against every grain of what I learned through WhatsApp if you're new here I was the 19s engineer at whatsapp back then we kept our scaling efforts really private we didn't really talk about our monthly active users publicly if your product is really genuinely good why would you bombard people with marketing videos instead of just releasing your product is just not how an engineer would usually operate there are already multiple videos dissecting every second of the marketing footage so I'm not going to delve into too much detail but the gist is that cognition posted a series of marketing videos and first off none of these demos were live they were pre-recorded and carefully edited okay now this one I do want to talk about a little bit can autonomously find and fix B bugs in codebase Deon helps Andrew maintain and debug his open-source competitive programming book one thing I want to highlight is how this person is using Deon for debugging but if you really look at the proms they're using incorrect extraction of Base powers in log class regular person looks at this prom do you even know what incorrect extraction of Base powers in log class even means anden hang who recently suggested that in the future we won't need to know programming but that's really not entirely true because so far AI is more like a calculator right if you don't know how to do math how would you know what inputs to give to the calculator like how would you know you have to use add or multiply or division so similarly if you don't know programming how would you know what to ask Devon like another backstory I can tell you I personally don't find it really enjoyable to collaborate with non-tech people because articulating technical issues is difficult for example I was building a website with a product person when she asked me for changes she would say stuff like the website looks a little off and I'm like what about it and she's like the font is a little off it was really frustrating because like what the is off right like do you mean you want more padding a different color the alignment like what exactly is off you were to tell Devon hey Devon something seems a little off can you fix it here Devon would likely be confused and this is not about anything against non-technical people but it's just really tough to convey technical specifications and prompts when someone doesn't really understand software engineering number five don't trust everything you see on Twitter the lesson here is to take Trends with the grain of salt I don't think AI is about to swoop in and take away all of our jobs anytime soon at least not in the immediate future in my previous video I did talk about exponential growth so I don't mean you should ignore AI altogether and assume your job is completely secure absolutely not but what I do mean is don't be so easily persuaded by marketing friends I've seen people panicking and contemplating dropping out of computer science just because of Debon or just because Jensen said that programming is going to be useless in the future so please don't do that honestly I don't really see a better alternative in terms of Industry to join anyways it's not like other Industries are really going to be driving anyways in the age of AI so if you want to learn more about the top five Rising roles and top five roles that are going to be disappearing in the future with the changes of AI check out this video otherwise this is the video that YouTube thinks you should watch watch next so I'll see you there