Over the last five years I've been in and out of therapy and 2/3 of my therapists have "graduated me" at some point in time, stating that their practice didn't see permanent therapy as a good solution. I don't think all therapists view it this way.
But it's not fully self driving. SF Waymo can't bring you to the airport. You missed OPs point, which was that the last few percentage points are the hardest.
They recently got approval for the airport, and the issue was legal / regulatory, not technical. They could have been doing rides from the airport years ago.
Okay how about this situation that one of my junior devs hit recently:
Coding in an obj oriented language in an enormous code base (big tech). Junior dev is making a new class and they start it off with LLM generation. LLM adds in three separate abstract classes to the inheritance structure, for a total of seven inherited classes. Each of these inherited classes ultimately comes with several required classes that are trivial to add but end up requiring another hundred lines of code, mostly boilerplate.
Tell me how you, without knowing the code base, get the LLM to not add these classes? Our language model is already trained on our code base, and it just so happens that these are the most common classes a new class tends to inherit. Junior dev doesn't know that the classes should only be used in specific instances.
Sure, you could go line by line and say "what does this inherited class do, do I need it?" and actually, the dev did that. It cut down the inherited classes from three to two, but missed two of them because it didn't understand on a product side why they weren't needed.
Fast forward a year, these abstract classes are still inherited, no one knows why or how because there's no comprehension but we want to refactor the model.
True True,
I remember another example, with Linus Torvalds, who at a conference used a trivial example of simplifying functions, as to why hes good at what he does, or what makes a good lead developer in general.
It went something along the lines of.
"Well we have this starting function which clearly can solve the task at hand. Its something 99 developers would be happy with, but I can't help but see that if we just reformulate it into a do-while instead we now can omit the checks here and here, almost cutting it in half."
Now obviously it doesn't suffice as real-world example but, when scaled up, is a great view at what waste can accumulate at the macro level. I would say the ability to do this is tied to a survival instinct, one which, undoubtedly will be touted as something that'll be put in the 'next-iteration' of the model. Its not strictly something I think that can be trained to be achievable though, as in pattern matching, but its clearly not achievable yet as in your example from above.
> Tell me how you, without knowing the code base, get the LLM to not add these classes?
Stop talking to it like a chatbot.
Draft, in your editor, the best contract-of-work you can as if you were writing one on behalf of NASA to ensure the lowest bidder makes the minimum viable product without cutting corners.
---
Goal: Do X.
Sub-goal 1: Do Y.
Sub-goal 2: Do Z.
Requirements:
1. Solve the problem at hand in a direct manner with a concrete implementation instead of an architectural one.
2. Do not emit abstract classes.
3. Stop work and explain if the aforementioned requirements cannot be met.
---
For the record: Yes, I'm serious. Outsourcing work is neither easy nor fun.
Every time I see something like this, I wonder what kind of programmers actually do this. For the kinds of code that I write (specific to my domain and generates real value), describing "X", "Y", and "Z" is a very non-trivial task.
If doing those is easy, then I would assume that the software isn't that novel in the first place. Maybe get something COTS
I've been coding for 25 years. It is easier for me to describe what I need in code than it is to do so in English. May as well just write it.
20 here, mostly in C; mixture of systems programming and embedded work.
My only experience with vibe-coding is when working under a time-crunch very far outside of my domain of expertise, e.g., building non-transformer-based LLMs in Python.
I mean, unless you just don't know how to program, I struggle to see what value the LLM is providing. By the time you've broken it down enough for the LLM, you might as well just write the code yourself.
Curious about the mechanics here — when you say the model was ‘trained on our code base’, was that an actual fine-tune of the weights (e.g. LoRA/adapter or full SFT), or more of a retrieval/indexing setup where the model sees code snippets at inference? Always interested in how teams distinguish between the two.
Sure I did. I'm aware his giving isn't just mosquito nets. That doesn't mean I believe the money is being directed correctly.
If your position is "it's his money so none of us should comment", I'd expect equal pushback on people saying "wow I really agree with how he's spending it."
Private industries are not incentivized to pursue the same kind of research as the public sector. An excellent example of this is antibiotics (or antifungal drugs). Pharma companies do not put forth a large budget for research like this, because it isn't profitable. But, if we were to see a super bug crop up in the next 10 years (unfortunately this is not a crazy "if" due to trends in antibiotic resistance), then you would quickly see the public sector's research paying enormous dividends in terms of missed economic hardship. These economic benefits would not affect the private sector in the same way, because a pharma company would not see, let's say, $50b in profits from helping ease an equivalent $50b in economic hardship caused by hospital bills/suffering/lost productivity.
On top of this, let's just look at the current private sector and what they spend R&D dollars in: can you really say that $1b spent by Google, Meta, Amazon etc. actually ends up being better worthwhile than $1b spent by NASA? see this list of inventions NASA has inadvertently created: https://www.jpl.nasa.gov/infographics/20-inventions-we-would...
In general, public sector research is already so strapped for cash. Their budgets are not large. Their salaries are lower than the private sector. I agree that public R&D spending is likely not at a "theoretical optimal level", but would you argue that private sector R&D spending is?
Second this, I know several close friends who have done the Recurse center and they say it's a phenomenal experience, one of the best things they've ever done. Generally speaking the job placement afterwards is also not all that bad if you decide to go back into the workforce.
This article feels a little bit pointless to me. ryandvm said it perfectly - a staff engineer is just what happens when you have a senior engineer who deserves a promotion, but wants to be a technical leader not a people leader. How that actually fits into the company varies heavily company to company, specifically size and maturity.
I feel like this article wants to talk about the performance matrix through the lens of four pillars, and just arbitrarily chose "staff" as the role to talk about since it's a very senior engineer.
I don't think any dem is saying that. I think they're saying that inflation is the result of a global pandemic, not Dems printing another 1.5T. I think by all accounts the economic landing after a global pandemic was really good, certainly better than 2008. We aren't in a recession. Prices are high, but so is employment and job growth. The government failed at something, whatever that something was: was it a failure in signaling that yeah, these times are hard but guess what, it's because of COVID and buckle up because we did the best we could? Or was it that they let inflation rise too high? I'm not sure.
What we’ve learned is that a politician should definitely not pull the lever in the trolley problem. Let four die instead of one, then claim credit for the one.
