That is a fantastic presentation. It reminds me of the annotated source/site that backbonejs used to have, but this one is interactive. Really great job!
Elderly gen-x are screen addicts because they had access to gaming consoles and computers? Maybe I’m misunderstanding your comment. Not every gen-xer was strapped to a console growing up.
My boomer parents and their friends are all staring to their phones wa-a-ay more often than I’d consider healthy. At least as much as my millenial/xennial friends.
Social media and attention stealing algos are addictive and unhealthy, regardless of the age group. If anything I’d say that gen-x is uniquely positioned - old enough to have experienced the world without the internet, young enough to see the consequences of it.
> but from time to time, doing this does require doing arithmetic correctly (to correctly add two exponents or whatever). so it would be nice to be able to trust that.
It sounds weird, but try writing your problem in LaTeX - I don’t know why, I’ve found a couple models to be incredibly capable at solving mathematical problems if you write them in LaTeX.
I’d add that the potential to support other architectures is also a benefit. At the moment the framework 13 chassis supports risc-v [0] [1] with rumors about an arm variant.
Beyond practical repairability and sustainability, I appreciate the possibility of swapping out a mainboard for another with a completely different arch
That jibes with what Nadella said in an interview not too long ago. Essentially, SaaS apps disappear entirely as LLMs interface directly with the underlying data store. The unspoken implication being that software as we understand it goes away as people interface with LLMs directly rather than ~~computers~~ software at all.
I kind of expect that from someone heading a company that appears to have sold-the-farm in an AI gamble. It’s interesting to see a similar viewpoint here (all biases considered)
> The criticisms in the cnn article are already out date in many instances.
Which ones, specifically? I’m genuinely curious. The ones about “[an] unfalsifiable disease-free utopia”? The one from a labor economist basically equating Amodei’s high-unemployment/strong economy claims to pure fantasy? The fact that nothing Amodei said was cited or is substantiated in any meaningful way? Maybe the one where she points out that Amodei is fundamentally a sales guy, and that Anthropic is making the rounds saying scary stuff just after they released a new model - a techbro marketing push?
I like anthropic. They make a great product. Shame about their CEO - just another techbro pumping his scheme.
Ditto - no spam at all. I do get reminders that it’s time to file, and conveniently, if nothing has changed since the previous year (address, employer, marital status, etc.) then they can pre-populate the form.
I’ve used them for probably the better part of a decade. I’ve never had any issues, they’re pretty inexpensive, they e-file by default and you get a notification when state and fed are accepted. The wizard is easy to follow and explanations are available. They have bells and whistles like audit protection, that seem reasonable but that I’ve largely avoided.
The name might seem a little sketchy, but it’s a legit service. I don’t know about “free” - I’ve always ended up paying to file at least one of state or federal, I don’t recall, but not pricy.
(Edit - looking closer, I’ve been using them since 2007. Wow, time flies. Great service, no complaints.)
That doesn’t seem very likely from my reading. It sounds like it’s not unheard of for prion-diseases like Creutzfeldt-Jakob to take a while to destroy a person. The poor lady fta apparently had a genetic component that contributed to the latency. Brutal. Inducing dormancy in this and similar cases would effectively present a treatment, if not a cure, to an hereto untreatable, uncurable disease.
> … the woman first visited doctors with tremors and trouble balancing. As is often the case, once symptoms started, her condition rapidly worsened. She was hospitalized … and several days into her stay, she fell into a coma that she would never awaken from.
In other news, apparently hgh used to be extracted from cadavers. Wtlf.
A “complex” cfg is still a cfg, and, giving credence to Chomsky’s hierarchy, remains computationally less complex than natural, context sensitive, grammars. Even a complex cfg can be parsed by a relatively simple program in ways that context-sensitive grammars cannot.
My understanding is that context sensitive grammars _can_ allow for recursive structures that are beyond cfgs, which is precisely why they sit below csgs in terms of computational complexity.
I don’t agree or disagree that LLMs might be, or are, capable of parsing (i.e., perception in Chomsky’s terms, or, arguably, “understanding” in any sense). But that they can learn the grammar of a “complex cfg” isn’t a convincing argument for the reasons you indicate.
> [Harper Reed] cautioned against being overly precious about the value of deeply understanding one’s code, which is no longer necessary to ensure that it works.
That just strikes me as an odd thing to say. I’m convinced that this is the dividing line between today’s software engineers and tomorrow’s AI engineers (in whatever form that takes - prompt, vibe, etc.) Reed’s statement feels very much like a justification of “if it compiles, ship it!”
> “It would be crazy if in an auto factory people were measuring to make sure every angle is correct,” he said, since machines now do the work. “It’s not as important as when it was group of ten people pounding out the metal.”
Except that the machines doing that work aren’t regularly hallucinating angles, spurious welding joints, etc.
Why? I can understand the argument that you don’t want an ISP or a middlebox injecting ads or scripts (valid I think even if I’ve never encountered it to my knowledge), but otherwise you’re publishing content intended for the world. There’s presumably nothing especially sensitive that you need to hide on the wire.
I’m seeing a lot about price hikes, new service charges, and bad customer service since the acquisition. I noticed the price hike with one of my domains through them, I paid it but was close to reconsidering. Thinking I’ll try transferring it now.
How is it winner-take-all? Training models is expensive, yeah, but big companies are willing to do it, and if deepseek is any lesson, it’s not as expensive as we used to think. You might be building a castle in someone else’s kingdom, but you’ve got a few kingdom’s to choose from. None have a moat and none seem to be going away any time soon.
It’s cheaper to move from one model to another than it is to train a general purpose model yourself (to say nothing of domain-specific smaller models or anything open source.)
I’m not sure about problems turning into black boxes, LLMs are pretty explicit in my experience when producing a solution (good or bad.) _How_ they came about that solution _is_ a black box, but that’s not a new problem.
Temperature? Definitely tweaks the results…but I don’t know if “deterministic” is a term you can use in any way, shape, or form, in the context of LLMs.
> It feels disingenuous but even companies I know and respect are adding questionable “agent” features that rarely work.
Word. I’m not a huge Microsoft fan, but it feels like they’re just shoving ChatGPT/copilot down your throat every chance they get. It’s integrated into everything - it’s useful when it’s useful, okay, but it generally isn’t. It’s one more Microsoft’ism that you have to learn to tolerate or simply ignore.
I can’t tell if Nadella’s really betting the farm on it all, or if he’s just trying to leave his mark.
I especially like the idea of getting a framework version in this case I want to swap in a different mainboard. By their own admission, the risc-v board is targeting developers and not ready for prime time. Also coming from the US, not sure how the tariff thing will workout…
That sounds a bit unfair. All that code that we neither wrote nor understood, I think in the case of Rust, it’s either the borrow checker or the compiler itself doing something it does best - i.e., “defining memory safety away”. If that’s the case, then labeling such tooling and language-enforced memory safety mechanisms as “a bunch of additional code…you didn’t write and…don’t understand” appears somewhat inaccurate, no?
> Monopolies can only be created by government dictate, like old AT&T or modern cable companies.
By virtually every definition I can find, a monopoly is a an entity that functions as the sole, or effectively the primary, provider of a good or service in some market. That seems to perfectly describe Google’s position wrt web-based advertising. Do other ad-platforms exist? Absolutely. Do they exhibit the kind of market dominance or control that Google does? Nowhere close.
> Google has no power to stop them.
Fact? I’d argue that Google’s sheer size and dominance means they don’t need to stop them. Potential competitors simply don’t stand a chance given Google’s size, number of resources, and reach. Explain how that’s not a significant factor into Google “power to stop” a potential rival?