Yes, in fact the military has a system in place to make sure that they only that they only use alternate suppliers and on Anthropic AI, They do this by placing a designation on Anthropic AI that it is a risk to the supply chain, which forces purchasers to use other suppliers.
I don't think this is how preferred suppliers have typically worked through the supply chain risk considering the US has definitely had preferred suppliers before and has not historically used this against a US company that was not on the list and the power of such a designation extends far past the military into all parts of the government. I'm pretty sure this has more traditionally been relegated to simple requirements or specifications.
Most of these models, both open and closed, are now so over tuned to agentic and coding tasks that they no longer work well for general purpose. Kimi K3 is an exception to that – maybe you need that larger size todo well on a broader range of tasks.
I've read it since 1995 and I vividly remember them enthusiastically stanning Bjorn Lomborg's denialism. That's when I learned that the Economist is actually the world's first manifestation of a phenomenon now associated with ChatGPT: immaculately-phrased ideological assertions that are presented as common sense that all good, clever, right-thinking people already know.
Like ChatGPT, the findings are delivered with polish and confidence, and like ChatGPT, you would not like what's in the sausage, were you to understand its provenance.
It's devastatingly effective on a certain kind of man.
Yes! This is something I've been thinking about quite a bit the past few weeks. We are going in this direction at work and I think graph storage is a natural way to think about this.
For my own edification, what is the goal of SPARQL compared to normal SQL or if you live in predicate-land, Prolog or Datalog?
The semantic web tools never seemed to have their breakout moment and seem to run in parallel to "mainstream" history of relational and logic programming, but I've never really understood why.
I started thinking more deeply about how I would actually do this today and it won't be as straight forward as I initially thought. Will keep it on the radar and see if I can figure out a clean way to implement. Great suggestions!
It seems wrong that Amazon should get in trouble for this when it was XAI that began this new race to the bottom.
Peoplecheered xAI on and admired how quickly they were able to move, without much regard to the methods or legality. IMO competitors had no choice but to respond in kind when they see XAI not just getting away with it, but also getting a pass on legal issues.
> It seems wrong that Amazon should get in trouble for this when it was XAI that began this new race to the bottom.
Who gives a shit who started it? The ones who continue engaging in it, should clearly be punished as they're working against humanity's best interest of keeping this planet survivable.
If you're one of the largest companies in the world, constantly brag about how environmental friendly you are and how much you care about "sustainability" or what the trendy word is today, and you end up building a data center that sets a record in pollution (either directly or indirectly), you're doing something massively fucked up and are misaligned with the rest of the people on this planet.
> Peoplecheered xAI on and admired how quickly they were able to move, without much regard to the methods or legality. IMO competitors had no choice but to respond in kind when they see XAI not just getting away with it, but also getting a pass on legal issues.
I guess this is what late-stage capitalism gives you ultimately, companies who feel like it's fine doing whatever because competitors are also doing whatever. No one I know have been cheering for anything like that, I suggest if you have people cheering about stuff like that near you, get far away.
I was with you until there. Qwen and the OpenAI models are great, aggressive agents, but they’re not as good as the anthropic models for human interaction. They just don’t have the subtlety, understanding, or attention to detail.
Claude 4.5/4.6 - absolutely agree. Fable 5? From my (limited) testing, also reasonable to interact with.
Opus 4.7/4.8/5? Absolutely smug and antagonistic and preachy. I'm constantly fighting with it to stop fighting me and accept that I occasionally know better. It's really frustrating to spend so many tokens of such an expensive model arguing with it.
yeah, it's honestly amazing just how badly anthropic managed to screw up something in claude after 4.6. its night and day, and every time I see claude doing something stupid, i instantly realize I was accidentally on Opus 5
Really? I've heard so many other people complain about this recently. And maybe it's possible that it's the prompt style even. But interesting that it's not across the board.
Pretty much! It's not dynamic memory hotplug or resizing, just dynamic billing. Usage is based on the actual RSS used by the app so you can think of the VM's defined size as just a limit, for both CPU and RAM. This means you don't have to spend time profiling and right-sizing instances to optimize costs, just make your application fast and efficient and don't worry about it!
The minimum fix would be to read it once per user turn, rather than once per tool call turn. As it is, if you tell an agent to update the file and it decides to do multiple edits on separate turns, you take a full cache miss per edit. If you're waiting 5-10 minutes on each miss (deep into a window, which is when you'd usually want to update notes) then you notice the difference.
The post is pretty explicit about it happening on every SSE turn, but I realise if I write something that is (to put it mildly) a rant, then I shouldn't expect people to read it charitably.
I also wouldn't mind being prompted before invalidating so I can decide whether to keep my cache, like most editors will prompt you if a file is edited externally.
Um, the 'codex world' is the OpenAI world and there is a ton of drama and product confusion there!
Anthropic has certainly had some drama inflicted on them by the US administration, but otherwise they have just had heads down and executed with great focus. That is why they have succeeded.
No, they want to have rules on how the military used Anthropic AI.
Military would still be free to use OpenAI for the nasty stuff
reply