I can’t stand all this hype anymore. They’re pushing it so hard that I’m starting to develop an AI phobia. If it were really that good, they’d already be making big money off it. But it’s not — it just needs your data (codebase?), supposedly to make it better. The fact is, at least they’ve gotten halfway there, which from their perspective, is sweet. All the AI labs are actually losing a ton of $$$ to this AI race, but the data, ohh yeah! Let's go, baby!
I am strictly talking about the coding capabilities of the LLM, and not their core LLM capabilities, which they genuinely excel at.
I'm pretty pragmatic and resist AI hype. However, I am in the camp that the tools we've built do not maximize the potential of the LLMs we have today.
I see 2 things happening in parallel.
1) Tools on existing LLMs continue to improve (cursor -> claude code).
2) The LLMs themselves improve which makes existing tooling better and results in new tooling to take advantage of the improvements.
I'm not sure when I see either of these slowing down and they've been accelerating at a very rapid pace. Perhaps when the funding dries up.
I think that's the question but I believe if we don't have any more LLM improvements that we still have a couple years of tooling improvements using what's there today.
I'm sort of surprised that coding is a leading use case but do not see any reason it would not spread to other industries (what the OP is saying).
I think this is not true in most of the cases. The (security) technology behind the debit/credit cards using the SmartCard chip (IC) is pretty ubiquitous. It is the same as the security technology guarding the SIM cards in your phone and even your eSIM. Basically the protocols and the interface specifications are the same. In the end, they are just smart cards. Imagine this technology not being strong enough, because I remember the days when the security of the pre-paid public phone cards was quite rabish and any kid with some skills and knowledge could forge a card with unlimited credit.
It very happens that the father of the smart card technology to be a french guy [1] and the current biggest provider of this technology is the french aero-space/defense/security company Thales Group[2] followed by another frech company called IDEMIA.
There is a very nice biography of the technology [3].
That summarization includes a video I hadn’t seen before where Calhoun explains the results in his own words and mentions the case of the Ik tribe which I didn’t know about:
> [00:01:55] at least one human group, the Ik tribe discovered in Africa by dr. Colin Turnbull (https://en.wikipedia.org/wiki/Ik_people#The_Mountain_People) has gone through a similar terrible process according to dr. Calhoun: the two conditions that they went through was that there was a decision on the then colonial government to make a wildlife preserve of this big valley which was their home as they are hunter-gatherers, they moved them up onto the mountainside and gathered together in much higher densities, they had been spread out small groups, all placed in one place and also required to engage entirely different way of life, which their social roles, the whole… everything they had learned how to do had no meaning. Under those conditions the social behavior broke down. Entirely they began living… they still live together but their activities was just directed towards themselves, everything you would think is human disappeared.
Probably this paper [1] is very well known among the classic HN audience, but I dare to leave this link here for the ones who missed it. It is an easy read and it just explains with plain words the backbone of the UNIX system as it was envisioned 50 years ago.
I use ProtonVPN and I guess they aren't that stupid to mess up the future of their company by abusing the customers of the service that started the company in the first place. I guess other VPN services offered by prestigious companies, like BitDefender are quite safe from this standpoint. Yes, the free VPN services have always a large risk associated with.
It's not necessarily a problem to use a service, (Signal?), which is funded by an intelligence agency. It can serve their purposes while serving yours too. If the service gets wide enough adoption, then the huge volume of traffic becomes a perfect place to hide their own intelligence activity.
I see this is the first and only post you have on HN. Since you shared with us the problem you have with Stripe, would you be so kind to share the address of your webstore so that we can make an idea about the stuff you are selling.
It is becoming hilariously unpolite to see all these complains revealing only half of the story.
Submitted a bug report regarding poor performance like 2 years ago and they are still "looking" into it. I guess the shiny look has priority over performance for them.
They definitely don't seem to care about performance or bugs very much, but it should be noted that the latest version starts up much faster even from cold storage. I used to look at the splash screen for ~8-10 seconds, but not anymore. It was the first significant performance improvement in a very long time.
Other than that, I agree with you. In addition to well-known performance issues and longstanding bugs, I'm also still waiting for features like full Wayland support that was promised by the end of 2020. I mean, I understand it's not on their highest priority, but why promise things you're not actively working on?
I'd suspect that it is a combination of the following:
a) a lot of existing code that needs to be untangled (e.g. moving various tasks to not depend on indices);
b) figuring out how to avoid unnecessary re-indexing (e.g. the downoadable index caches for things like Java files);
c) having to define a migration path to spread out breaking changes to give plugins enough time to adapt.
IIUC, some of the performance work with the latest version is due to parallelizing the various background processes, which uses coroutines, so you could add "implementing language, library, and tooling support to Kotlin/Java" to that list.
This is rather a simplistic view with the wrong conclusion of what the education means to our society and culture. The scientific progress has always been carried out by a small percent of population, and not a mass process. Its only recent when the education was made available to the masses.
I also suspect the author mistakes the term education for kwoledge. And yes, some knowledge is lost, but that is due the technological progress. It's always been that way.
I am strictly talking about the coding capabilities of the LLM, and not their core LLM capabilities, which they genuinely excel at.
reply