I don't mind the name change because it more correctly signals what the position has always been -- a secretary of war with a "secretary of defense" mask on.
And then one day, Astral creates a new Python implementation in Rust or something that is way faster and all the rage, but does this particular thing different than CPython. Whoops, you can’t use that runtime, because you now have cursed parts in your codebase that produce nondeterministic behaviour you can’t really find a reason for.
and then all the serverless platforms will start using Astral's new rust-based runtime to reduce cold starts, and in theory it's identical, except half of packages now don't work and it's very hard to anticipate which ones will and will not and behold! You have achieved Deno
In the definition of RAG discussed here, that means the workflow looks something like this (simplified for brevity): When you send your query to the server, it will first normalise the words, then convert them to vectors, or embeddings, using an embedding model (there are also plain stochastic mechanisms to do this, but today most people mean a purpose-built LLM). An embedding is essentially an array of numeric coordinates in a huge-dimensional space, so [1, 2.522, …, -0.119].
It can now use that to search a database of arbitrary documents with pre-generated embeddings of their own. This usually happens during inserting them to the database, and follows the same process as your search query above, so every record in the database has its own, discrete set of embeddings to be queried during searches.
The important part here is that you now don’t have to compare strings anymore (like looking for occurrences of the word "fanfiction" in the title and content), but instead you can perform arbitrary mathematical operations to compare query embeddings to stored embeddings: 1 is closer to 3 than 7, and in the same way, fanfiction is closer to romance than it is to biography. Now, if you rank documents by that proximity and take the top 10 or so, you end up with the documents most similar to your query, and thus the most relevant.
That is the R in RAG; the A as in Augmentation happens when, before forwarding the search query to an LLM, you also add all results that came back from your vector database with a prefix like "the following records may be relevant to answer the users request", and that brings us to G like Generation, since the LLM now responds to the question aided by a limited set of relevant entries from a database, which should allow it to yield very relevant responses.
Do you have anything to support that claim? Carbon taxes are a theoretically effective mechanism to tilt the markets towards more sustainable means of production, that is something most economists agree on; alas, practically they are often thwarted by caving out exceptions or delays for short-term political gain.
It's an ugly and wasteful system set up instead of other, simpler measure that were politically unacceptable at the time, like higher VAT, excise duties on all fossil fuels across all industries without exception, including fuel oil for heating and aviation fuel.
Mmm. The language is not precise enough - if most economists agree on something it probably is true. If the corporate media gives the impression all economists agree on something, it is probably not true.
Economists as a profession understand extremely well that they have no ability predict the economic future beyond what the futures markets say.
There are great standard CMSes that do everything technically better than Wordpress (not that it's harder to jump higher than a rock, but hey). That's not the hard part. Every developer should build a good CMS once.
The hard part is displacing Wordpress market share; building a community of bloggers, marketeers, agencies, web designers, and so on; creating a huge ecosystem of paid and free plugins, allowing plugin devs to commit to your marketplace and lock customers in.
Wordpress is awful. The only thing it's got going is its moat, but that's not an engineering problem, but a people problem instead.
Those standard CMSs that are technically "better" than WP, I would bet they are at least over a decade old, some have even come out during the same time with wp when there was no market share to speak of and still were left in the dust.
The problem is that people misunderstand why WP was and is better than all alternatives that tried to take it's place, I have no idea either but I know that others have tried same thing as CF and failed.
I have been one of the maintainers of a moderately successful CMS back in the day, but there are definitely well-groomed alternatives that have popped up a lot more recently than that, especially from the headless variety.
I'm convinced the thing that WP did better was being the first simple and accessible blogging platform when blogging was still a thing. IIRC, the alternatives were things like Joomla or Drupal - awful behemoths for enterprise users. WordPress was a breath of fresh air compared to those, and out of the blogging scene, people started to use it for agency projects, while others published an ever-increasing number of plugins and themes. The rest is just momentum of that movement.
I find it hard to believe that people used to WordPress, with its flaws and virtues(yes, wordpress have virtues), will switch to this, no matter how much it's from Cloudflare.
CraftCMS for example is a great example of what you can achieve with PHP; there's Directus or Strapi for headless CMS, for example; or Ghost, Contentful, Storyblok and more as SaaS alternatives.
I tend to disagree. When guiding AI through many rounds of code review, it can self correct if shown where general issues exist. It does take practice for using the language of the model, i.e. drift instead of issues. Human in the loop is good enough to produce useful and accurate code today.
If you can actually do this, please sell your services. You will become a multi-millionaire overnight if you can provide a workflow that doesn't result in mass hallucinations or incorrect suggestions you're able to do something no other LLM company can.
The more common use case is that these tools struggle immensely on anything outside the happy path.
reply