yes, and theres no evidence that they arent (or can't) use profitable inference to subsidise those other expenses. Some companies will keep spending massively to train better models, and some other companies will not, and offer good api prices. Which will end up being used? That depends on whether the spending turns into better value models
Imo its pretty clear that anyone who is taking the issue at least somewhat seriously knows the amount of value they provide is not non-zero. However, the problems are manifold: firstly, toolchains vary wildly, from fancy autocomplete, to engineers chatting with codebases they're unfamiliar with, to people integrating them into devops and infra, to people doing spec driven development, with a thousand philosophies inbetween. Many people suspect that those above them in the ladder are on the cusp of massive failure due to losing track of the code, and many people higher on the ladder think those below them are overly cautious. I hate to be the guy saying "oh it must be somewhere in the middle", but I will say at the very least I like being able to use it to read docs for me, and to synthesize syntax and simple scripts (give me a join that works across these tables and gives me column x, y and z - give me a python script that parses a file like this example and extracts abc data - given this api spec figure out how I can get this data from this endpoint, go)
as for building actually complex software, the art of that is not in simply chaining together such scripts. Its the art of using architecture and testing to shape uncertainty, and developing requirements (and extrapolating sensibly from incomplete requirements). I don't think llms are great at this, but they arent terrible either. A lot of the more active users in the space are doing stuff where theyve realised they need more detailed specs, which like, yeah, we knew this already - better defined problems lead to better software.
I agree the most interesting use cases I've heard of are about increasing the rigor of software development practices, but there's definitely a lack of coherence in methodology.. I believe that some users and companies are successful in this effort, but the odd (and interesting!) thing is that so far we don't seem to know how to communicate how to do it successfully.
what's wrong with artists having jobs via a program? whats wrong with struggling alcoholics having jobs via a program? athletes? politicians? there is no inherent virtue in the struggle and effort associated with great mathematical achievement. It may be satisfying and worthwhile for the solver, but not for society at large, any more than any other pleasurable activity. No, as it is, the sole reason for it is in the result itself. In increased understanding, as it flows down into the sciences, and engineering. There are other benefits, recreation and joy as experienced by others, from access to beautiful proofs, though these are never explicit goals of such programs because they are both impossible to quantify and rarely ever remotely relevant compared to the value brought by the practical value brought by maths.
Of course, there may be some valid arguments that everyone should have a jobs program in the form of ubi or something similar. But I feel thats very different to arguing for mathematicians specifically
for mathematicians, they do a form of fundamental research that is
1. (generally) incredibly cheap to fund, and
2. (occasionally) has extremely out-sized commercial impacts.
This is to say that jobs programs for math (and more generally fundamental research) have lead to extremely positive ROI for society, which is the typical justification given for funding them.
it arguably still is. The primary unit of production of the jobs of mathematicians is itself not particularly useful for society. In this sense funding them is a jobs program. It is also true that they occasionally produce things of great value, and more frequently the things they produce can be leveraged by other researchers to directly produce things of value. But neither of these are what the job of a mathematician is (either in a day-to-day sense, or even for many mathematician's careers).
To go back to the analogy of jobs programs for alcoholics, it is somewhat similar if there was a small chance every time an alcoholic defecated in public gold came out. This fact might be used to support a jobs program for alcoholics, on the basis of it being positive ROI to society. At the same time, the "job" any individual alcoholic is doing in this setup is not particularly useful to society, so one might still call it a jobs program.
its not an article, its a github issue. There is no proof, someone posted a screenshot of another person on twitter complaining about a vague bug they experienced. It's certainly not clear the maintainer used an llm in the way described (how would you know unless they told you?). Its not even clear what issues there are specifically, or whether they were caused by ai usage.
yep. Its not even capitalised in the title of the linked article. Honestly though, "wiz" is such a weird word to use, have they turned into the Sun magazine? they use "boffins" too in the article...
it seems to me this is a ratcheting effect. If a larger than average is discouraged, even implicitly by being called to a meeting, people will target under the average. Repeat this and you'll quickly find the average plummeting
not to mention the other english word, "missed", which sounds exactly the same.
However, people really do have a great deal of trouble with this exact concept. Its a common test to give to people to define a function with a misleading name, like multiplyByTwo(int x, int y), and then show them the code inside (return x/y)
and ask what this results in. Surprisingly many non programmers have a great deal of trouble with the idea that the name can be different to the behavior, especially in cases where its subtly different not entirely like this.
Other fields have similar issues, where people have trouble with this sort of abstract thinking, divorcing abstract knowledge about a thing from its appearence. iirc its a stage in child psychology and development, and shows up in teenage years after puberty. Many adult people never develop it fully, and struggle with this to a greater or lesser extent
as technologists I think it's tempting for us to look at the idea of the economy, or at least like the idea of a planned economy, in terms of an optimization problem. Ordering the most "deserving" inputs and outputs, those that lead to some level of human flourishing, over whatever timescale and via whatever primary, secondary, tertiary effects etc and assigning resources to them in proportion to their importance - and that the problems that need to be solved are identifying what those best orderings are, and logistics.
I think this is something of a mistake. The best use of resources isn't nearly the biggest problem facing a planned economy. The biggest problem, currently seen clearly in both russia and ukraine simultaneously, is that when a planned economy crops up, for war production or socialist reasons, there is a great temptation for those involved to do corruption, at whatever level they can. falsify reports, change production numbers, all sorts of crimes to accumulate power over the flows of resources, such that that power can be leveraged to trade for other things.
The real problem to be solved is a social one. How to make the watchers who watch the watchers? Corruption can in principle never be fully prevented entirely - enough people believing that corruption goes unpunished and any system will fall. But building a system convincingly self-sustaining enough that everyone can't quite be sure that others aren't checking their work (and in a more positive framing, one where people feel good work is both expected and rewarded), one where people can be confident corruption will be found, will pay dividends in people policing their own behavior.
I agree with you that this is not primarily a logistics but really a social problem. I think the section on the Soviet Union in the article addresses specifically that point.
> The proponents of the economic calculation problem now tell us that the proof of the failure of socialism lies in the collapse of the USSR. However, what failed in the USSR was not the planned economy, but the lack of democracy.
> The economy was not democratically planned by the workers, but by a state bureaucracy. The Stalinist bureaucracy usurped the democratic power of the workers and peasants who had led the 1917 October Revolution and expropriated the capitalists from power. The misery and scarcity created by the Civil War and the invasion of 21 imperialist countries, including Canada, was the manure on which the bureaucracy grew like a weed.
> In the 1920s and 1930s, the planned economy allowed the USSR to move out of its semi-feudal conditions and developed the means of production sufficiently to defeat the Nazis and become the second world power. But from the 1960s onwards, production became more complex, and the bureaucracy gradually became unable to calculate all the information to plan production properly.
> Indeed, how could it be possible that a few hundred civil servants in Moscow could plan an economy of hundreds of millions of people? The bureaucrats wanted to meet their production quotas, even if it meant cutting back on product quality and lying about the economic information in their departments.
> What authoritarianism and lack of democracy in planning does is to degrade economic information and undermine planning. This is what explains why the USSR has entered into economic stagnation. Bureaucrats then brutally re-established capitalism in the USSR in 1991, leading to shortages, mass unemployment, the return of prostitution, and more.
> As Trotsky said, “the planned economy needs democracy as the human body needs oxygen.”
An example of what that democracy could look like is presented in the next secion as well.
> There are several ways of imagining the democratic mechanisms under socialism. It will be up to the workers themselves to decide them. But, for example, we could have general assemblies in the workplace, so that the workers can decide together how to organize that workplace. We could elect workplace management committees, recallable at any time and accountable to the workers. Workplace and neighbourhood assemblies could send delegates to city and regional committees, to a national congress, and eventually a world congress, which would be responsible for coordinating the entire supply chains from production to distribution. Scientists, technicians and engineers could advise the assemblies and committees in their decision-making.
> State-of-the-art computers and algorithms from nationalized Amazon would be used to assess changes in demand. Information would be sent directly to workers in the collectivized factories so that they could coordinate storage in warehouses and distribution centres. These computers will be able to calculate production costs and the number of hours of work that go into the production of each commodity. And this will help elected committees to plan product prices based on the needs of the population.
> If your requirements are reasonable and serve the needs of end users and the business
Have you ever heard of the idea of malicious compliance? doing exactly what is required, and no more, is the path to ruin for all of us. Requirements will never work this way.
reply