As someone who has been doing this since the mid 80's in all kinds of enterprise environments, I am finding that the latest generation are getting rather good at code like that, on par with mid-senior level in that way. They are also very capable of discussing architecture approaches with an encyclopaedic knowledge, although humans contribute meaningfully by drawing connections and analogies, and are needed to lead the conversation and make decisions.
What LLM's are still weak at is holding a very large context for an extended period (which is why you can see failures in the areas you mentioned if not properly handled e.g. explicitly discussed, often as separate passes). Humans are better at compressing that information and retaining it over a period. LLM's are also more eager and less defensive coders. That means they need to be kept on a tight leash and drip fed single changes which get backed out each time they fail - so very bright junior in that way. For example, I'm sometimes finding that they are too eager to refactor as they go and spit out env vars to make it more production like, when the task in hand is to get basic and simple first pass working code for later refinement.
I'm highly bullish on their capabilities as a force multiplier, but highly bearish on them becoming self-driving (for anything complex at least).
> I'm highly bullish on their capabilities as a force multiplier, but highly bearish on them becoming self-driving (for anything complex at least).
Very well summed and this is my exact stance, it's just that I am not seeing much of the "force multiplier" thing just yet. Happy to be proven wrong, but last time I checked (August 2024) I didn't get almost anything. Might be related to the fact that I don't do throwaway code, and I need to iterate on it.
Recently used Cursor/Claude sonnet to port ~30k lines of EOL Livescript/Hyperscript to Typescript/JSX in less than 2 weeks. That would have took at least several months otherwise. Definitively a force multiplier, for this kind of repetitional work.
Do not know how you are using them?
It speeded up my development around 8-10x, things i wouldnt have done earlier i'm doing now, e let it do by the AI; writing boilerplate etc. Just great!
I just used Claude recently. Helped me with an obscure library and with the hell that is JWT + OAuth through the big vendors. Definitely saved me a few hours and I am grateful, but those cases are rare.
Amusingly, lately I did something similar, though I still believe my manual code is better in Elixir. :)
I'm open to LLMs being a productivity enabler. Recently I started occasionally using them for that as well -- sparingly. I more prefer to use them for taking shortcuts when I work with libraries whose docs are lacking.
...But I did get annoyed at the "programming as a profession is soon dead!" people. I do agree with most other takes on LLMs, however.
That sounds like they need a Unix plug-in pipe approach creating small modules that do one thing well handing their result without caring where it goes to the next not caring where it came from module while the mother Llm overseas only all the Blackbox connection pipes with the complex end result as a collateral divine conception.
While developing an AI tool to help hiring managers prepare for interviews, we stumbled upon what seems to be a novel method for detecting bias in Large Language Models.
By comparing how LLMs (Claude, GPT-4, Gemini, Llama) interpret anonymized vs. non-anonymized versions of the same content, we can measure and quantify bias reduction. The interesting part is that this technique could potentially be used to audit bias in any LLM-based application, not just recruitment.
Some key findings:
- Different LLMs show varying levels of bias reduction with anonymization
- Llama 3.1 showed consistently lower bias levels
- GPT-4 performed better in specific tasks like interview question generation
We're a boutique AI consultancy, and this research emerged from our work on building practical AI tools. Happy to discuss the technical implementation, methodology, or real-world applications.
Luxury goods prices are largely related to the cost of production in the form of quality rather than any other form of scarcity. The question they answer is 'what is the very best product possible?' Luxury leather bags for example have a huge degree of craftmanship (which is why fakes are mostly lousy and very distinguishable). This is why they become a well-known brand in the first place - it's a mark of quality in design and production.
Rimowa is an example where they have had to constantly innovate in order to maintain a quality edge. Service is another area - if you buy a Hermes silk tie, you can get it cleaned and pressed for life for free at any Hermes shop in the world (cleaning and pressing silk ties is incredibly hard and they often get destroyed by dry cleaners).
The problem occurs when the brand itself becomes the cash cow as a result of this consistency in quality, and businesses start to exploit this. They will usually slap their brand on an outsourced production line product and charge a premium. Occasionally that might be justified by the product having a great distinguishing design to compensate but often it's not the case - and that's where we start to think of it as a bit of a scam. An example is Church's Shoes. From 1873, they were around the highest quality brogues in the world available at any price (short of bespoke shoes) and were handmade in workshops in Northampton. In the 1990s Prada bought them and started transferring production of all but a top end line to Italy, where the shoe manufacturing was partly automated and inferior. Unless you are buying the handmade bench grade shoes, you are now paying extra for the name and probably being a bit of a mug unless you are very attached to the particular design. Now that brand is tarnished for the high end shoe market and dependent on consumer ignorance for the rest. (As a sidenote, a lot of the craftspeople who worked for Church's before the takeover moved to the other Northampton shoe makers who used to be considered lesser quality and some of those brands are slowly moving into the luxury end as a result of the quality improvements.)
I will very happily buy a nobrand or knockoff where there is no functional difference or the price tradeoff doesn't justify the difference, but often top end goods can justify their price.
Could be the timing was wrong? I'd think that an online direct to consumer model could work now with focus on social media marketing. Worst case you get bought out by De Beers if you make headway, especially now they are getting in the consumer synthetic market themselves.
It's just the path to commodity that every product follows. That's inevitable with gems now that synthetic stones are equivalent, so the choice facing De Beers is to stick with expensive mined stones and die out while the market evolves, or to get a big chunk of the action. The only thing is that to pull it off they had better be ready to cannibalise the existing business and let it go. They need a very long term plan which doesn't always revolve around the next few quarters which might be a tough sell with shareholders.
Drupal 8, plus integrated ElasticSearch (so you don't have to build UIs and can also integrate search into displaying collections generally), and Drupal Commerce if you like.
Have you worked much with D8 Commerce? I know it's spoken of highly in the community but after tolerating it since my first deployment of that combo in late 2015, it still feels half baked at best and almost completely undocumented relative to other major Drupal projects.
Why this very evening I'm having to manually re-input authorize.net details because they updated the module with a very poor upgrade path.
Well it's always been under resourced and you could fairly say it lacks product/commercial maturity so will have many rough edges and gotchas, but at the same time it's a very complex problem space where all of the solutions have drawbacks - so I'd look more at what the differentiators are, since we have to endure pain in any case.
So for example, with shopify you know it's all going to be click and build and largely configurable by commerce owners. Ease of use and a generic set of capabilities that scale across its userbase.
With Drupal Commerce, you know you are going to have a completely extensible open source system which leverages very powerful existing Drupal components such as entities, views, users, rules, search. There isn't really anything else like this.
A long time ago, I saw one very big d6 project (with insanely huge spiky traffic) which recognised this and used d7 commerce as a separate standalone system which sat on the back end integrated with the d6 site. This was after assessing Hybris, Magento, and Demandware as alternatives and understanding that they would require far more bespoke coding and have higher maintenance costs to come anywhere close to the very very specific and high spec requirements this particular site had.
The fundamental principle of presumption of innocence means you shouldn't go around spying on everyone looking for signs of criminality.
And the idea that the tech is more dispassionate and therefore fairer than humans is a step on the road to hell. I guess you never had an anomalous credit or insurance rating yet. (Notably GDPR has a clause for precisely the ability to review automated decision making where it can have a material effect.)
