Discuss HN: Software Careers Post ChatGPT+

I've been using ChatGPT all weekend to generate code and what I found was this:

  * Its absurdly good at coding and following types. For example if you change a type in Rust to be an Option, it will refactor the code to properly handle Options in the parts it used them. but it isn't perfect.
  * It gets it well enough there. It can generate test cases so its easy to test to see if it works.

  * it introduced many many many subtle errors. Like variables not being initialized or not splitting the children correctly.
  * Its implementation worked when all keys inserted were in order, but not when they were out of order.
  * it would miss and leave out variables.
  * It would frequently have index errors from trying to access lists out of bounds.
  * writing the code in Rust was almost impossible. it would constantly have wrong types or move errors.

Even the test cases that it generates can be deceptive. They look convincing but upon closer inspection, sometimes are just not really testing anything.

> Or is it all bogus (with justification), and we're fine?

This is mostly my take. We are at the stage of truck drivers ten years ago who might have been freaking out about self-driving taking their jobs, and here we are in 2022 with a truck driver shortage.

I don't think GPT can be useful to the point of replacing software engineering without a consistent mental model of itself, time, and the world, and I predict it will approach the limits of what advanced-search pattern-matching can do without getting anywhere near those AGI capabilities.

I mentioned it in a previous comment. It's common for senior engineers not to write that much code. They spend time their time on meetings, planning, creating architectures, presenting solutions, discussing solutions, triaging, keeping up-to-date with tech, clarifying business cost, working on waste avoidance, reviewing code, streamlining processes, vetting new tech/solutions and, in general, understanding everything that is going on.

ChatGPT+ will definitely have some affect on junior devs but us more experience folk should be fine... for now..

> Are lawyers at risk?

There are a lot of sub-areas of expertise and practice that someone with a JD might choose to specialize in. I have some small personal experience in (technically) advising a failed/defunct startup that sought to solve the problem of patent search using an AI. This was years ago, now...maybe around 2018ish. The endeavor failed for various reasons, but it did provide some insight that's relevant to your question here.

As these language models become more advanced (and much more accurate) I think there will be a number of ways in which they will disrupt existing domains of human expertise. Note that I used the word disrupt and not displace. In the patent search space that I was lightly involved in for a short time I basically learned how expensive and time-consuming a good patent search actually is. The machines were planned to be leveraged to drastically reduce the time and cost of typical prior art searches, but would still require human touch-points to interpret results and make final decisions/reports. I think this sort of use-case is much more in line with what will ultimately happen in any sort of foreseeable future. The AI will supplement and ease the previously-human-only task. It will not supplant/replace it.

Here's the problem-- this will automate the type of "programming" that's just looking something up on stackoverflow and more or less copy/pasting the answer. There's a lot of that out there.

Once you're a more senior engineer there's a lot more than just writing code. Designing a system, worrying about maintainability, operational burden, scaling, etc. are where you might spend your time more.

I'd argue that even for "programming" the usefulness is debatable. These models spit out relatively correct code but mainly in the sense that they regurgitate something akin to a SO answer. There are plenty of subtle logical errors though and that debugging exercise often takes longer than just writing the code. Lots of code references other libraries, etc. and APIs do change frequently. So ensuring what's generated even works as expected is a fair amount of effort.

Still, the chasm between "engineering" work and "programming" work is only going to get bigger as a result of tooling like this. I expect a lot of what's currently outsourced to overseas IT consultancies can be replaced with half the staff leveraging these tools. The bottleneck has always been producing the exact requirements, tightly scoped tasks, etc. though. We're no closer here.

Let's say you have an idea for a mobile app, but you have never coded anything and you don't want to learn Android/iOS programming to make this app. You want to convert your idea into an actual app that is deployed on the app store where people can download it on their phone. What would you do? Normally you'd hire a programmer and get them to build, test and deploy this app. Now you have ChatGPT, are you going to use it to build and deploy this app? No. You will still hire a programmer who will probably use ChatGPT to write some or most of the code as opposed to writing all the code manually. This will save them a lot of time and since they can complete the project faster, they will charge a lesser amount to you, it may still be higher on a per hour basis but it's a win-win for both you and the programmer.

AI assisted coding or anything else will not replace the professionals who had been doing it manually so far. It will only make them more productive. They can do more work in less time and charge more for the enhanced productivity.

I really fear for the day I need to debug some AI-generated legacy code. It's not really the algorithmic part that scares me, but the naming and code architecture.

These AI seem so confident when they output BS that it makes you doubt yourself. Now imagine if some code looks coherent but you find that each line does something slightly different that what the variable names and other method calls suggest. Now you can't trust the names to build a mental image of the code; you have to follow each method call to find out exactly what it does. It would be worst than looking at obfuscated names because you may think you know what is going on.

A perspective from a non-professional who has been teaching himself to code:

My knowledge of exact functions is poor. I might know that I can use Framer library to animate on-page elements, but I have little to no understanding of the exact function needed to animate an object from, say, left-to-right on hover.

My normal workflow was to either read the documentation or search StackOverflow for answers. I would then have to rework the function to fit my current use case.

Now, I've been asking chatGPT directly to build the exact function for me.

So far, it's been a massive timesaver. I'll probably learn more if I dig through the documentation, but since I'm a hobbyist, not a professional, it's much more convenient for me to just get the information I need, without digging through Stackoverflow or documentation.

We all have repositories with open issues. I'd suggest that you open up github, click the first issue you have open, and try to close it with ChatGPT assistance. For me, this revealed a gap between its spectacular skill at programming puzzles and its total incompetence at closing issues- it can do dynamic programming challenges instantly but can't modify (a function that loads a file from disk, or downloads that file if it isn't already present) to (the same function but it correctly handles a failed download)

There’s a bias towards coding capabilities amongst testers (and perhaps trainers) of ChatGPT.

I bet it has (or can have) a similar level of capability when it comes to chemistry, biology, finance, law, etc. - all human knowledge that is expressible as text or formulae.

With regards to software engineering, a great amount of time is currently being wasted by coders trying to solve problems that have already been solved by others. Some of the solutions for that previously included libraries, SaaS, and Stack Overflow.

Now we have another tool at our disposal to 10x all software engineers (and perhaps the same for other industries).

The problem with artificial writing is writing alone doesn't accomplish much. Art is the expression itself, so AI art is the final product. But with writing, the goal is to often communicate something novel, in which case the AI wouldn't know, or transact and get something done, in which case the AI won't have the authority or access.

Take customer service chatbots. It really does not matter how "human" the bot is, if it isn't given any power, such as refunding an order. It can't really apologize either, without lying about its lack of empathy, creating a real life turing test between it and the customer.

Of course, businesses justify making refunds hard, so they often will prevent even humans that power, and arm them with templates and legal excuses anyway. Which explains why most customer service is sh*t. They're not there to help you, as much as they are paid to entertain the thought of helping you.

With code, at least the code is itself the end result, like the art. And it can be tested with execution and with unit testing. But most programming is still deciding what to name functions and what functions are needed. The implementation is just the technical task that "anyone" or "anybot" might do. It's the mindless part.

So like AI driving, the last mile of the problem may prove to be close to impossibly hard. Or at least, it isn't an AI problem that can be solved with current AI. It needs another breakthrough. Until then, even AI will continue to rely on human input. The furthest we'll get with AI coders is AI coders taking instructions from a coder, with the end result heavily dependent on what the human adds, not what the AI adds.

As a developer, I legit feel crippling anxiety and might need professional help.

People saying it's not intelligent, really don't get it. It looks like it's working exactly like the human brain. We, humans, also pretend to know what we're talking about and often talk with authority without really understanding what's going on. When you apply for a react.js position you pretend to be a js God when in fact you know shit, just enough to be able to stitch things together to make them work.

The rate at which technology, languages and frameworks change also doesn't give you the ability to master your craft. In fact, mastering a specific tool/framework is risky for your career because of new tools emerging every 3-5 years. Every company uses a variation of react,vue,python,go,java,ruby whatever, it's impossible for the human brain to keep up and be good at everything. We can only scratch the surface of the complicated mess of a tech stack and that's exactly what gpt3 is extremely good at right now. So yeah, I fear a lot...

I can't imagine how good this is gonna be in 5-10 years. I legit feel scared and anxious. How can we prepare for this future?

None of these AI systems significantly outperform searching for the answer online. The AI is good because: - The programs it generates are small - The programs generated solve well known problems with well known answer

Current AI replaces SW developers the same way StackOverflow and blog posts replace SW developers.

The real question is how fast new AI is going to be developed which overcomes these challenges. I think what is pretty certain that current language models will never be a serious contender against programmers, their whole structure makes it very difficult to perform actual logical reasoning about the structure of a computer program.

Another thought: How useful is an AI which is correct 99% of the time? How much development resources would be needed to fix a code base which is 1% total nonsense that does not even compile or is just plain wrong. Would the developer time (all of which have never seen the code) spent fixing that code base be actually less than a total rewrite?

IMO for now it is the next iteration of stack overflow/google. I will 100% use this to ask a question on how to do something at the API/function level because it gives really good results.

Ex: "How can I create a webserver in python" will generate all of the boilerplate instead of me going and looking at the flask documentation.

> any layperson can say "hey AI slave, write me a program that..."?

I think this is one of the big challenges. From my limited experience, a person without expertise being able to describe accurately enough what a program should do is often a challenge. Further more it is often after an initial implementation that even engineers are able to understand the full requirements of a system. It's possible that generating yet-another-crud-app wouldn't be any challenge to GPT but then at that point there could be a market saturation of CRUD applications (is that not the case already?) and value might shift to apps that are able to do more than just CRUD.

Then again, predicting the future is tricky business when we are so notorious at not being able to predict what can we do during the next sprint! :-D

I feel like this is a false opposition of AI vs Programmers. I would argue its never going to happen IMHO because AI + programmers is going to be so much better than either on their own.

Even if you have some thing generating amazing code you still need the energy to deploy it and check that it works and put the docs in the correct folder and email somebody.

So IMHO the future we all need to be looking it is how these tools can make us all more productive and then the juniors just get a different bunch of shit to learn. Maybe we develop more focus on writing tests to "prove" the software is correct as opposed to reviewing the code output? Idk, but I don't see the industry suddenly replacing devs given that the problems are not entirely reduced to simply writing code.

I do think its worth wondering about, BUT

* Most non trivial systems are complex, thousands of lines, many systems working together, solving complex domain-specific problems and not toy examples we can produce from ChatGPT

* CoPilot and ChatGPT introduce subtle bugs, so to use thes for code generation, you'd have to write many tests. A LOT of tests to confirm it worked. At a minimum this requires careful thought

* Getting the prompt to do very subtle things right seems time consuming. Why not just write the code (perhaps with guidance from AI)?

Maybe its a fun way to get started, and more convenient than a search engine, but I'm not convinced YET you can make it do anything non trivial without a lot of work and conversation with ChatGPT.

ChatGPT feels like the current aim assist debates in a lot of FPSses to me. It'll make you better at the shooting part of the game, perfect even. But, won't necessarily make you that much of a better player, because aiming is only one aspect of what makes someone good at FPSes. However, if someone is generally good enough or very good at the "not aiming" portion of the games, then having aim assist would drastically increase their overall skill.

Also, ChatGPT doesn't really work for any UI based programs at all. For example, if you want help using Excel, you'll get a list of instructions, but nothing visual to help you out.

I feel AI tools will just accelerate software development speed overall rather than replacing software developers them entirely. A lot of boilerplate code, how to configure X and the 80% of development side will be taken care by these AI tools remaining 20% will be left for the software developers to fine-tune and making the functionalities work together.

It's like code completion on steroids, still lacks depth and detail. It can create code mashup and glue it perfectly in no time, so it's worth a lot as most of the dev work looks like that. So I predict that it will be a standard tool at some point. Still, it will require a skilled dev that understands what to ask and how to interpret the answers. If somebody would ask me if I agree to be operated by a "junior" surgeon who is helped by such "AI" I would said no.

It will raise the bar for software development significantly.

GPT needs training data. Once a problem is "figured out" and there is enough information about it to reach a treshold where it's learnable by the AI, it will stop being something done manually.

There is reason to believe that most commodity software (e-commerce, content management, media asset management, any CRUD-with-makeup, infrastructure automation, etc) will be commoditized even further to the point that no software developer would be needed to create a product that leverages these approaches, or any reasonble combination of them.

Sure, there will be a need to guide the AI to the best solution. A need to harness it properly. Some software development skills will be needed for that, and my guess is that this will become a temporary occupation in the near future, either explicitly or not.

Eventually, there will be problems that large models can't solve. Not even partially. Stuff that has no training data and there is no way to acquire inferred knowledge about it. We'll know what that is only when the commoditization of what we already know is almost complete. My guess is that it will be something trivial but rarely considered worthy exploring, writing about or pursuing, but valuable. Some kind of impossible miracle software that will be not that impossible in the future (complete protocol/data portability, wide compatible standards across thousands of different systems, etc).

Or we all will have to learn how to flip burgers in a boring dystopia. That's a reasonable scenario as well.

It is a personalized StackOverflow on steroids, so I expect it not only increase productivity of boilerplate tasks, but also teach how to program in the process.

In addition to classically trained, self-taught, we're going to have AI-taught programmers. There will be some curious cultural differences there.

We're the old guard now: "Listen kiddo, in my days, we had to google for stuff! We were lucky if there was any documentation to read, manually!"

I've been looking at simple mathematical logic problems that are computer science adjacent and while I've had a little success on some simple toy scenarios, I've found it incredibly difficult to get it to provide even "directionally correct" solutions for anything useful.

Coding is probably a pretty restricted problem space. After all, there's only so many ways you can handle a string until something throws an error.

The way I look at it is it is an excellent companion that will help me automate a lot of tasks

1. Send this quick email to my team telling I'll be late. 2. Adding boilerplate stuff that I keep doing again and again. 3. Writing very quick and dirty poc stuff 4. Do simple usage analysis of logs/metrics for learning some patterns

After these, the onus would be on me to focus on what needs to be done.

Until the "AI Slave" is able to formally identify that it has produced a correct solution, which requires the ability to identify and correct non-trivial mistakes including at the "conceptual" level, and probably to provide the "rationale" for its "thinking", it is unlikely to affect this particular area too much.

oh wait until you realize that most software is basically the same crud and most "unique special solutions" to problems are not so unique as you have thought.

Finally the most common reason why the "stack is hard" in startups is because there was an "all knowing" coder who build a system which he cant be replaced out of.

As a freelancer is always nice to see how quickly they shift their mannors if you show them that "you can do it too" and that all that is needed to break a stack like that is reading up on the 15 frameworks they used to build it.

All in all i dont think AI will replace us, but mainly because its going to be cheaper to hire a studied IT veteran than to use the AI.

Are the times where we are treated specially over? yes indeed. We never where more than handyman for computers.

I think we will continue to be surprised by how effective GPT-X and other LLM models are at coding tasks. The main reason is that unlike math programming is a field where dozens or hundreds of correct solutions exist for every task.

All an LLM has to do is generate one out of hundreds of potential programs which solve the problem. It doesn't even have to generate perfect code, lots of profitable businesses are built off of buggy software.

What's wild about this is that for now ChatGPT seems to have the most benefit for 'stack overflow' programmers who rely on copy-pasting solutions to problems other people wrote. Now they can effectively generate a stack overflow answer to their specific question instantly!

For now AI seems like it will empower them.

I tried getting ChatGPT to output me a bilinear tangent steering guidance for a rocket yesterday and it never quite gave me what I wanted (although it seems to have regurgitated some interesting stuff that it had been trained on which was kind of neighboring to what I wanted--I think it was pulling from autonomous vehicle trajectory optimization rather than rockets).

It was pretty good at simple stuff like taking one steering algorithm I had and decomposing it from a directional vector into pitch/yaw and doing the substitution for me. It does high school algebra pretty well, easier than using symbolic computations in Matlab.

As I have said elsewhere many times. You still have time to make it big with crypto or startups. If you engineer your way into high frequency trading you can also make a lot of money. If you can do this and FIRE you can become an early investor in the companies that will domanince the mutlipolar ai era in the years approaching the agi level. Do this correctly and you will eternally stand apart from the ubi pleb class that many of our less competent brethren no doubt will join.

Response by ChatGPT:

It is certainly possible that in the future, language models like GPT-3 could become advanced enough to compete with human programmers. However, it is important to note that these models are still limited in their capabilities and are not yet at the level of human programmers.

At present, offshore outsourced coders do not pose a significant threat to the software industry, and it is unlikely that language models like GPT-3 will be able to fully replace human programmers anytime soon. However, it is possible that these models could be used to automate certain tasks and make the work of human programmers more efficient.

If language models were to become advanced enough to compete with human programmers, it is likely that software engineers would need to adapt and develop new skills in order to remain competitive. This could include focusing on areas where language models are not yet advanced, such as visual and fine motor skills.

As for protecting themselves from potential automation, software engineers can stay up-to-date on the latest technology and continue to develop their skills in order to remain competitive in the job market. It is also important for the industry as a whole to advocate for policies that support workers and ensure that they are not displaced by technology.

Overall, while it is possible that language models could become advanced enough to compete with human programmers in the future, this is not currently the case. Human programmers still have a valuable role to play in the software industry, and it is important for them to continue to develop their skills in order to remain competitive.

I haven’t seen any code generated by an AI that in any way approach the kind of complexity I am dealing with in my daily job (large scale enterprise software running operations and planning for international companies). Especially given the fact that just describing what the software does in English to the AI and then testing it and fixing the mistakes made would probably be as much work as just coding it in a programming language. And we already have a technology that gives us instant highly advanced well functioning code we can reuse in our work: It’s called frameworks/modules/libraries/… Programming is not like art. We are already directly reusing millions of lines of code written by other programmers. AI’s for programming will only be interesting when it comes up with truly new ground breaking algorithms that humans haven’t been able to come up with already.

ChatGPT provides incorrect solutions to most iOS prompts relating to UI stuff that I gave it.

They would look correct to someone who knows how to program, but does not know iOS. Most of them even compile. But they're very wrong, even when I tried dozens of different inputs in an attempt to get the correct solution from it.

> or writing "Hello World" in Haskell

It's already writing Haskell [1]

[1] https://twitter.com/runarorama/status/1598940278190395392?s=...

If you are a programmer you will know that debugging your own code and other people's code makes up a significant portion of your job. The GPT coding systems I've seen so far don't make much of a dent with that aspect of a programmer's life.

I think for now, it behaves mostly like a Junior SE. While the code is decent - 1. it has errors sometimes, hence doesn't compile 2. it's not the most optimized code 3. it could have potential security vulnerabilities, since it doesn't have access to internet and hence isn't aware of the latest exploits particular to an industry or a programming language (guess the models used for training aren't the most up to date)

So, all in all, we are ok for now, as long as we keep levelling up ourselves, and evolve beyond script kiddies.

But hey, we never know what could happen in the next 10-20 years if it continues evolving at this pace.

Has ChatGPT been able to solve problems that are novel and unique in software engineering? For example, if you are a SE in a ninche domain, would this understand the domain specific language needed to code in that environment?

A bit of a different perspective, but ChatGPT could help people prepare for Leetcode style puzzles very well. SEs who don't practice these type of puzzles often can benefit from ChatGPT if they intend to prepare for Leetcode style interviews.

I tried using ChatGPT to debug hard Leetcode questions and it was able to debug its own wrong solution together with me and I personally learned something from it.

https://twitter.com/willtai_/status/1599565171051331585?t=Ve...

It's probably too early to tell, but my wild guess is that it will add another conceptual level of abstraction to programming. Instead of say pesudocode→code→assembly, you would get to description/case→code (via GPT)→assembly. I put assembly there on purpose; just like higher-level languages made programming more accessible and opened up the profession, I would expect this to (eventually) happen here as well. Now, how do we get there from here and which companies lose out in the new world, who knows? we hadn't had a paradigm shift like this since the iPhone.

The obvious answer is that deciding what to build is difficult, but that rather invites asking whether building lots of stuff at random and throwing it at the wall would have a better hit rate than building the right things.

Interesting angle is testing where successfully tying a "real program" to one of these AI tools would look a lot like a fuzz tester. Possibly a better one than the current state of the art.

Philosophically I'm not hugely worried about automating the wisdom of Wikipedia and Reddit for mass application. Seems more likely to rot than improve itself. But we'll see.

Can I ask what are the hopes of having a truly open source/self hostable version of chatgpt? I really worry about the implications of giving a single company so much control over the development process

Personally I don't think our jobs are at risk, but perhaps some junior-level positions might have a reduction in volume.

More than anything, I see this type of system being used in combination with a system like Github Copilot to produce much more useful suggestions through feedback loop.

Things that used to be long, annoying-but-easy tasks may be offloaded to ML.

Remember that GPT is ultimately a very fancy Markov chain. It does not and cannot reason. It's more likely to replace busywork and give engineers better time to work on important architectural features.

I think the programming working model will change as it did years ago when we moved from punch cards to files and then from assembly to higher-level languages. Programmers will continue to program solutions with assistance from code automation for less business-focused parts of the problem at hand. Consider the more analog version of this: nowadays we spend more and more time copying answers from StackOverflow than we do debugging at a lower level.

Until it can change existing code, I think we're ok.

cf. Englebart; and Ashby "Introduction to Cybernetics"

The fundamental principle of intelligence augmentation (of human intelligence by machine assistance) comes from selection: the machine can generate options beyond those which the un-augmented human mind can reach, but from which it can select.

Anyway, the question is, "What is good?"

It's an open-ended intelligence test built into the structure of the Universe, of all possible Universes.

Additional references:

Wendell Berry "What Are People For?"

Larry Niven, the character Teela Brown (spoiler alert)

"Five Whys" diagnostic technique and "Core Transformation Process" a similar technique for human consciousness

Also "Intuitive Ordinal Notations"

https://github.com/semitrivial/IONs

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

Can we discuss teaching post ChatGPT+?

I'm considering giving my local high school system a call to raise awareness. When I was in college 10 years ago I abused Wolfram Alpha for some math work, but this is a whole different level and may require a new model of instruction.

Why give free response homework when an AI can just generate 80% of it?

This will radically alter homework as we know it :)

I can’t believe I have to say this, but please consider not trivializing slavery by referring to twiddling transistors as such.

I think documentation will be one of the first places we see it used a lot. I say this because today I'm writing documentation about how to use sprints, Jira and projects within our company today. I've written over 50 percent with the help of ChatGPT. Lets me focus on the less boring written stuff.

keep in mind the current iteration of ChatGPT doesn't try to execute the code, its understanding is purely based on "reading" existing code.

This tech could potentially be integrated with a compiler and trained through self-play like AlphaZero (possibly with human guidance at a meta level). They could also add some kind of visual modality for UI code.

Once this happens, 90% of devs that work on CRUD style apps will lose their jobs, imo. The software industry will probably become a lot more stratified, like the music industry - a small percentage of ML researchers and devs making a ton of money at the top, close to zero for everyone else.

Software development is most at risk vs other "soft" industries, because for the most part code either works or it doesn't, and it's fairly easy to check.

Could we try having humans learn to evaluate other humans without leetcoding?

I mean it sounds fucking crazy just typing it out on my keyboard, but could we possibly do it?

ChatGPT can and will make us way more productive.

The obvious conclusion is that you're gonna need less engineers overall and especially on the junior and mid level.

I think in-house counsel for contract reviews, etc are most at risk. These AIs can flag risks pretty well and have a real attorney triage issues.

as a computer science student i am slightly concerned but I also don't think that it would necessarily be possible for AI as it currently exists to be able to complete software engineering tasks with the required level of specification/efficiency. But maybe in 10 or 20 years.... ? who knows what the landscape will be like

> Are these language models better than current offshore outsourced coders? These can code too, sort of, and yet they don't threaten the software industry (much).

I felt kind of offended by this biased view (especially the “sort of" part). Just because one is not from the same country as the OP, it doesn’t mean they are any inferior.

Quality Assurance / Testing / Verification and Validation

Like all professionals , SEs should not rely on paycheck long term. Should build some form of wealth like real estate, or virtual real estate (an app or sth) to profit from. Don't think the crazy salaries of the '10s are repeatable

test

Here is an unedited transcript with OpenAI's GPT-3. My prompts are peppered throughout, but I believe them to be obvious.

/START

Please list some microservices I would have to create in order to implement a system to recommend a daily workout to a user. Assume the user's information is stored in a database.

1. User Profile Microservice: This microservice would be responsible for retrieving, updating, and storing user information (e.g. age, weight, height, etc.) from the database.

2. Workout Recommendation Microservice: This microservice would be responsible for recommending a daily workout to the user based on their profile information.

3. Workout Tracking Microservice: This microservice would be responsible for tracking the user's workout progress and updating the database accordingly.

4. Exercise Library Microservice: This microservice would be responsible for storing the exercises and other relevant information (e.g. descriptions, images, etc.) in the database.

5. Notifications Microservice: This microservice would be responsible for sending notifications (e.g. reminders, updates, etc.) to the user.

Please list a suitable tech stack for 2 people to implement these services with.

Front-end: HTML, CSS, JavaScript, React.js

Back-end: Node.js, Express.js, MongoDB

Testing: Jest, Enzyme

Deployment: Docker, AWS

How would you implement the Notifications Microservice using these? Be specific.

The Notifications Microservice could be implemented using Node.js and Express.js. Node.js would be used to create the API endpoints for the microservice, and Express.js would be used to create the routes and handle incoming requests. The microservice would also need to connect to the database in order to retrieve user information. Once the user information is retrieved, a notification could be sent using a library like Twilio or Sendgrid. Finally, the microservice would need to log the notification in the database.

Why did you choose twilio or sendgrid over FCM?

I chose Twilio or Sendgrid over FCM (Firebase Cloud Messaging) because they both provide a more robust set of features and capabilities than FCM. Twilio and Sendgrid both offer advanced features such as personalization, scheduling, segmentation, analytics, and more, allowing for more flexibility and customization when sending notifications. In addition, both are more reliable and have better uptime than FCM.

Can you define the api for the notifications service?

The Notifications Microservice would have two main API endpoints:

1. POST /notifications: This endpoint would be used to create and send a notification to a user. It would require the user’s ID, the content of the notification, and any other relevant data (e.g. expiration time, priority, etc.).

2. GET /notifications: This endpoint would be used to retrieve all notifications for a given user. It would require the user’s ID as a parameter and would return a list of notifications.

/END transcript

This is remarkable. Absolutely astounding.

The amount of paranoia and fear mongering over gpt is insane. Firstly you don't get paid to churn out lines of code. You get paid to build, maintain and own systems. Secondly GPT is impressive but it is in the end a statistical matching machine prone to constant errors. It's very good as a tool to help humans but by it's very nature it's a terrible tool for automation. Compilers, high level languages and infinite free public libraries are far better tools to automate programming than GPT is. If your job is something like taking an English language question from someone and writing a short script for them in a programming language then I might be worried. But for the vast majority of programmers that's not the case

Well it would appear that leet code is easily replaceable. Not even sure why people would get paid to rewrite algorithms that can easily be imported from existing libraries. But actual development will take a while to be replaced. A developer does much more than write code. And good developers already know they should use off the self solutions for repetitive tasks. So at most this type of tooling will help, but it wont be anywhere near what their evangelists and the media claims it to be.