Not op but a similar experience. I used cursor with claud to generate a small bash script to set up a Postgres container, iterate over local migration files, and apply them using psql. The generated code did all correct but called psql from within the container. Even when feeding the errors back in it could not correct the code or identify the bug.
I am curiouse, how complex was the app? I use cursor too and am very satisfied with it. It seem that is very good at code that must have been written so many times before (think react components, node.js REST api endpoints etc.) but it starts to fall of when moving into specific domains.
And for me that is the best case scenario, it takes away the part we have to code / solve already solved problems again and again so we can focus more on the other parts of software engineering beyond writing code.
At no point does the author suggest that AI is not going to happen or that it is not useful. He expresses frustration with marketing, false promises, pitching of superficial solutions for deep problems, and the usage of AI to replace meaningful human interactions. In short, the text is not about the technology itself.
In simple economics, a decrease in price typically results in an increase in demand, unless the demand is inelastic.
Anecdotal experience: Onset of tools such as NumPy, which made it more feasible for a wider range of people to write their own simulations due to drop in cost (time/complexity). This, in turn, increased the demand for tooling, infrastructure, optimisation, etc. and demand for software engineers increased.
Yes our jobs will change but there are way to many problems to be solve to assume demand will not increase.
If you define elite programmers in the context of actual coding as those who excel at implementing ideas and solutions, I could imagine that this skill might become less relevant with the advent of AI. Smashing out over 1000 lines of Haskell would then be the equivalent of being able to calculate complex numbers in your head.
However, if you define elite programmers as those who possess good domain knowledge, communication-, management-, and soft skills, then yes, they might become so productive that they could replace developers whose main skill is writing code as we move up a level of abstraction. While it might help today to have a certain level of understanding about Assembly and C, we do not need to be elite at it to be a good software engineer.
I am asking as I met a few devs who are electrical engs. with a very good understanding of how a computer actually works but now earn more with React and Python.
Would the increase in productivity not happen to almost all people if they were given the interesting work and the direct support of a good leader to help them focus? At least, I feel a difference like night and day when working on things I am interested in, as opposed to when I am doing chores. This includes thinking about the problem when I am on the toilet, during sports, on walks, on the bus, and so on.
I flew with them a few times as well as KLM, Emirates, and Singapore. Lufthansa's customer service, on-board experience (i.e. food & entertainment), is not on par with the others. When a flight got canceled last minute, we were left hanging with a number to call and a long waiting line, and we are still waiting for our reimbursement. On the need to reschedule due to a positive coronavirus test, it took 16 minutes with KLM to reschedule and almost 3 hours with Lufthansa and a bunch of (online) paperwork.
On top of that, some strange management decisions like completely stopping pilot training and laying off the ones that started during Covid while getting financial assistance from the state and 2 years later struggling with a pilot shortage.
My two cents: I worked in a different engineering field before transitioning to Software Engineering because "coding" was and is what we need to solve problems, and I got the hang of it. A few years in, I spend little of my day actually writing code but more time in meetings, consoles, documentation, logs, etc. Large language models (LLMs) help when writing code, but it's mostly about understanding the problem, domain, and your tools. When going back to my old area, I am excited about what a single person can do now and what will come, but I am also hitting walls fast. LLMs are great when you know what you are doing, but can be a trap if you don't and get worse and worse the more novel and niche you go.
That is right, but it is a global market where Novo Nordisk can decide to sell in a country or not, so I do not think they would sell below cost. I agree with the point that the US pays disproportionately much for the R&D portion of the cost, though.
How is it different from the normal market? If I have a product and sell in Germany for 50 USD and Switzerland for 100 USD because that is what people are willing to pay, would you also say that Switzerland pays for R&D while Germany gets it below cost?
Indeed, my mistake and agree. Here, I wanted to refer to the fact that the transfer of ownership could likely be done even without involving a notary or a human in a properly digitalized system.
There are ways to ensure the correct transfer of ownership without involving a third party. You can see these principles at work on some trading platforms already, be it for Magic cards or something else where parties cannot trust each other because they do not know each other.
Next, you would expect that the notary would educate participants and act as a source of trust, an actor in your best interest, but that is not the case. Notaries can change contracts until the last minute, and unless agreed upon, the common 14-day withdrawal period for contracts does not apply to things like buying property. Furthermore, if you are inexperienced, you can easily fall into traps.
As a concrete example, when buying a part of a shared property, it is commonly believed that the "Hausordnung" (house rules) is the owners' agreement for house rules. However, that is not the case, as there can be more, and in our case, it forbade us to keep dogs. Now, you could argue that we should have made ourselves more familiar with the law, and I would agree that is true. However, it begs the question, why do we need a notary?
I've used React over the years, but now I mainly use Svelte. I like React, but for me, Svelte feels more lean and straightforward. Depending on how Threlte (a 3D library for Svelte) develops, I might switch completely. Additionally, React's Server Side Components are now introducing a paradigm shift, so it might be the right time to make the switch.
I heard good things about Bishop however I am a SE that would like do know more about what the ML team is doing and maybe work on some ML side projects. Would you recommend Bishop here or is it considerer to theoretical for such a case?
That is true for a lot of western countries but we are expect to produce and consume more concrete until 2050 then we did so far in history in developing countries.
I am always a bit disappointed when only few arguments are provided. Ken Griffin might have a good understanding of the implications of LLM or not but that is hard to tell here.
Especially law will be interesting to watch. Lawmakers have to encode their intentions into text. LLM are good at detecting patterns in texts, find inconsistencies and so on. On top of that I would argue that we learn from the past to predict the future. Laws do not change as frequently as tech does. So LLMs might turn out to be excellent at understanding the law, at least written law, and experienced from all the cases they saw during training. I think law is in for a change similar to software.
It gives you a lot of freedom and doesn't force many rules like other non-dynamic languages do. This makes it great for trying out ideas and working with data and machine learning. However, when you're writing more complex programs, you need to be more careful. That's why I feel like I encounter more bugs during runtime compared to other languages. I mentioned Go because it makes you stick to a smaller number of ways to do things, which feels more "pythonic".
I think we push complexity forward. I agree in the sense that the pure dev part will require less bandwidth for most but the free bandwidth allows us to push complexity forward into different domains. My father still needed to punch card to code and now we can setup an app with a few clicks world wide that uses NN to solve a task and that all by ourself.
So demand will be high for cross domain knowledge like Fullstack/ML + Domain X.
Isn't that the nature of tech? In the past most programmers needed to focus on low level details while today most devs kit together libraries and services and yet there are more then ever and salaries are higher then ever.
I think nobody that enters tech expects that in 20 years we "code" as we do today but we will still build stuff and need to solve problems... and there are enough problems to solve.
A possibility, and I think the way we work will evolve as well. But coming from a more "numerical" background, I can imagine a different route. Twenty years or longer ago, people who wanted to process a larger amount of data needed to
understand low-level details, compilers, C/C++/Fortran, mathematical details, and so on. Today, we have JAX, scikit-learn, and many more tools. But these tools did not make the old 'numerical' people jobless; instead, their jobs evolved. Today, we have more data scientists than ever. You can create your own app faster than years ago, including hosting, persistent storage, load balancing, ... And again, we have more web developers than ever. The same goes for jobs like DevOps and other jobs that I probably don't even know. The level of abstraction got higher as you better now what the algorithm is doing but you do not need to implement it again. The point is the field will evolve, and right now, it may be the biggest jump ever, but that does not mean jobs will go away. We might end up in a situation like self-driving, where we are really close but still missing the last bit and need human intervention. I hope LLMs will solve the tasks that have been solved many times before, like bootstrapping a CRUD app, and we can focus on the edge cases and niche problems.
Don’t you think that web3 would also end up centralized?
The underlying technology might be decentralized but the ownership is not. For web2 we have more or less working mechanisms like antitrust offices, on the blockchain we do not. In a proof of stacks scenario we would have the power distributed in a plutocratic fashion without mechanism for the many to counter balance the mighty few.
It is so good to see progress in these areas. They are so fundamental that we do not think about them (fundamental in the
sense that we just assume they are there) a lot.
I have some infos on heat-pumps but do you have a good source about the state of smart-grid?
A lot of companies build web api's around their ML/AI code and data pipelines in Python, is Mojo suited for these task
as well or is it specialized for numerical/AI/ML tasked?
They put some work into the Python adoption and it is enough for smaller projects but I did not try it for to complex use cases. You can see it under examples.
What is nice is that you can set the option to parse into PyDantic rather the dataclasses.