defnotai
If current tools are incapable of generating code beyond a certain level of complexity with consistent accuracy, then how are they capable of accurately judging human output? This reads like a puff piece citing some person’s unreviewed, “Microsoft management in a black box” algorithm.
It says a lot that this brand is worth enough to someone who wanted to revive it, if for no other reason than to capitalize on its infamy.
Accept it and move on. Cultivate community in your life, let go of the ways that technology no longer serves you.
> If the argument is that Sheil is duplicating Gifford’s existence, there’s something to be said about the fact that the items both of them promote are also imitations of someone else’s work.
Work can be abused and addictive just like any other drug. Optimize life to fit your needs, but if every moment spent not working makes you feel like you should be working, you might want to look into that.
I have a friend who works in logistics for GE and they’re getting training on the basics of GenAI and then they have to go out and find ways to integrate it into their workflow. The problem is that management isn’t doing the legwork to understand how to integrate the tooling, they’re just handing that responsibility off to the actual users. Those people wind up complaining that taking time to integrate these tools winds up slowing them down and they struggle to find meaningful applications for the LLMs.
It’s like management is saying “here’s a new hammer, we don’t know how to use it but the guy who sold it to us convinced us you can figure out how to use it. So go out and do it and be better and faster at your job, good luck”.
So Nvidia produces the chips, invests heavily in “preferred partners” that buy their top line chips in bulk, and promises that those preferred partners don’t get any special treatment, yeah right.
This whole thing reads like the first act of a Michael Lewis book.
And then, like it or not, companies are hopping on the agent train hoping to automate out a percentage of their headcount because that’s how they’re being pitched on it behind closed doors.
I wanted to like this, but it fell really flat for me. He led with emotion and came across as bitter and angry, meandering around several examples and calling Facebook execs perverts. Those accusations could very well be true, but it always makes me question the legitimacy of the speaker because that’s such a common cheap shot in politics nowadays and it distracts from his main point.
I think we need more folks like Ed using their platforms to call out what they see because I think he’s right overall.
- Godwin’s AI Law
If this winds up going through, people all over the country are going to see the sticker shock at the store very quickly.
Putting this out there puts OpenAI on the clock to release their own alternative or adopt this, because otherwise they run the risk of engineering leaders telling their C-suite that Anthropic is making headway towards better frontier model integration and OpenAI is the costlier integration to maintain.
Even Apple Intelligence is getting a lot of negative feedback by the review crowd due to its limitations like being unable to summarize a very large document (which is pretty much the point of such a feature).
The problem is that AI has few well-defined use cases and a mountain of expectations, and this really shows in the execution by these companies. It’s hard to build good products when the requirements are “we don’t really know”
I think he was threatened by WPEngine’s approach as a more popular offering than Wordpress.com and decided to go scorched-earth on them instead of trying compete on the merits of their product. He’s threatening anyone that tries to build bespoke hosting with extra tools like ACF.
We have no reason to believe that AI costs won’t skyrocket long term either. It feels like the early days of streaming right now, but how long will it last?
Foundation models are pretty reasonably priced depending on the workload, but having a plurality or majority of companies all hooking up to the same two or three model APIs is gonna get really interesting when prices start to increase, because your exposure to the price increase scales with the number of times that same provider is used by your vendors, your vendor’s vendors, etc.
If Altman wakes up one day and decides to double the token price, it could be Armageddon in the SaaS space.
In my experience, semantic search systems are very hard to test and don’t always return great results when given a very small or vague query, which seems counter to how people are taught to think about search.
And then if you want to do a hybrid search with reranking, those network calls start to add up and degrade the UX.
And then each month, there’s a new post about some new architecture that promises another search accuracy or performance boost, but you’ll need to rebuild the entire search system to find out if it’s effective for your use case.