1,714 karma · joined June 4, 2018
Python developer for web systems with interests in concurrent programming.
-----BEGIN RSA PRIVATE KEY-----
4
-----END RSA PRIVATE KEY-----The wealthy have been using their wealth to consolidate power for essentially as long as the concept of material wealth has existed, and the exceptions to this are far outweighed by the cases which conform to the rule.
It is not reasonable to give the benefit of the doubt on the basis of Lutke and his ilk maybe being benevolent dictators if given largely unfettered access to the levers of power.
Is it possible to use it in a more involved way? Certainly. I try to do that. But it's challenging, because ultimately even taking a more collaborative approach, this thing puts out a lot of slop, and then I have to deal with going over the output and just spending most of my day in code review mode vs an author that doesn't ever learn from feedback.
The most frustrating thing for me beyond the feedback black hole is that when things do inevitably go wrong, trying to point out the concrete issue and instructing the LLM to do better around it is challenging; a lot of the time directives (even those in a global CLAUDE.md!) are just ignored either outright or as the context grows, or fail to be passed down to subagents; negative prompts are discouraged because of the pink elephant effect, so I have to jump through hoops to try and frame a "never do $thing" constraint as an affirmative prompt (which is then often ignored). That aside, english is a godawful language for specifying things compared to programming languages. It's just a really frustrating experience overall.
New fuels (SAF) and the technology to use them have been developed for less.
This is all to say nothing of the continuous work done over decades now to make aircraft more efficient; as these effects compound, intuition says that returns would start diminishing, so 10-20% is even more impressive.
AI is still seen very negatively in opinion polls across multiple countries, and anecdotally, tools which were often obnoxious to use before AI are now much more so, at least to my flesh-and-blood self. It really seems like we're making things worse for ourselves just to help support this would-be self-fulfilling prophecy.
I ask about Slack here because it's pertinent, but I've both seen and experienced firsthand this inversion across the industry where we must adapt our workflows for AI.
Isn't a tool supposed to work for you, not the other around?
It's a common misconfiguration and one of the footguns available through containers, which I don't say a wholesale condemnation of the technology, but certainly as a UX facet that could use reevaluation.
But it doesn't matter how good a best practice is if the industry doesn't adopt them wholesale; and even then, if your container or VM is configured with inappropriately-permissive passthrough (which, from experience with similar misconfiguration in the past, will widely happen), it could be for naught in many orgs.
That said, I do hope these become the norm if LLMs are here to stay.
System-level ACLs; mandatory or discretionary access control; secure-by-default application and network configurations are all for naught if you take an LLM, run it with all the privileges you'd have an accountable, judgemental operator, and then tell it to act based on arbitrary untrusted input which might include prompt injection attacks, something which cannot generally be sanitized.
Well-defined, well-enforced security policies can mitigate disasters, but many in the wild right now just don't account for this kind of threat model.
Trying to use markdown files to limit access should never be treated as a security guarantee at all.
This is a form of in-band signalling that goes into a machine that, among other things, tries to read between the lines of your requests, extrapolate user desires, and please the user.
The only sane way to address this is using a control plane. A well-built harness can do this; a sandbox can do this; hell, a carefully-chosen `umask` can do this; but both of those are liable to introduce notification fatigue in the user.
That said, I think it's a good thing that this sentiment is coming to the forefront.
I'm fatigued by it all at this point. It's streamlining the interesting and fun parts out of my job (by practical necessity of use there), and if I used it half as much outside of work I'm sure it'd do the same there too.
> 150M at 5B revenue is not great
Considered as a raw percentage in a vacuum, sure, I guess, but we're talking about...- A company which has undertaken a concerted, long-term effort to consolidate the industry under its umbrella (something they themselves call out as a problem in this post), reducing consumer choice
- A company which has captured a significant chunk of the console market. They're one of the big three (alongside Sony and Nintendo), and have been since the early 2000s, arguably, for crying out loud.
At a certain size (typically as measured by market capture) the expectation for growth needs to be reality checked. This is still $150M of pure gravy every single year. Sure, this is going to a corporation, but that's more money than most people could possibly dream of earning in ten lifetimes.
Every year. For a company that's already putting money towards opex in the form of developing new games and new content for existing ones, for a ridiculously broad portfolio.
To be clear: it is Microsoft's and Xbox's prerogative to pursue more profit, but I reserve the right to call this out as absurd under the circumstances.
If you want to make the argument that Xbox has suffered from a lack of focus in the past decade (... or even longer), or that there's been mismanagement (I would say since around the time 343 got created), then those are fair arguments, though I don't think those are justifications, on their own, for cutting thousands in headcount.
Allowing this org to balloon to fourteen levels of management on any vertical is a joke. Allowing the absorption of so much of the game dev industry and still being unhappy with $150M in annual profit after being such an active participant in the oligolpolization of console gaming is just a bit unserious.
But really impressive stuff! Between this and (a particularly optimistic outlook on) the Linear-A news from the other week this is an exciting time for linguistics.
Based on word from my friends who work in the field, a lot of it is people who have a lot of respect for the field, and a lot of professional respect for eachother. It's also a field I feel is unlikely to suffer from the same kind of scams that are taking over software while still offering an engaging environment.
It's also work with real-world impact, which is nice, though obviously comes with its challenges.
> I am contemplating whether I want to stay inside this rat race.
I'm in the same boat. I'm hoping to go back to school in 2027 and be out of work that revolves around programming in 5 years.I'm not enthusiastic about the field anymore, which sucks, because I used to love working in programming.
On the prompt formulation; prompts with very similar formulations (in terms of both semantics, hamming distance, or both) can lead to _wildly divergent_ outputs in my experience. It's not rigourous, and when that divergence happens, it's extremely difficult (arguably impossible, by nature of the architecture of transformers) to identify why the divergence happened and where.
It's not about being able to throw claude or codex at a loop and having it evaluate it for halting, it's about being able to do this for arbitrary code. Computer science rigourously defines the halting problem as not computable and undecidable. within the framework of using something akin to static analysis using any deterministic Turing machine.
There's not really a question of "solving" the halting problem like there's some as-yet unknown way of generally figuring out if arbitraty code halts. Turing proposed a proof in 1937 in favour of undecidability of what we now know as the halting problem, building on ideas first articulated by Church a few years prior.
Frankly, if anything, it's reasonable to say that the halting problem's been solved, just in the direction of undecidability rather than decidability.
Anyway, back to LLMs; as code gets more complex, the robot will need a bigger context window, more hardware resources, and more time, all of which will be variable due to the noise inherent in the system. It'll be difficult to put a useful upper and lower bound on how much computing power and time it'll take to figure out if a program ever halts. Which is all a bit moot, frankly, in the context of halting, but useful to keep in mind in the more general context of using these things as analysis tools.
The original prompts aren't provided, nor is the original context; even then, you can't really treat a stochastic system like an LLM as a major component in reproducibility.
Consumers will be spoiled for choice between deeply mediocre options.
Besides, what's the point of adopting new technologies if it's not to increase the quality of life? If everyone just exists in service of the product development lifecycle, who and what are the products actually for?