I can only assume this whole post is meant to be an ad for the cyber frost model? The charts being unreadable such that only cyber frost is identifiable, benchmarks being chosen to mostly support it, and the model being only 2 days old all makes me rather suspect.
I think they're suggesting that lock in implies apple didn't write the code to help support 3rd parties. Lock out implies they actively wrote code to prevent 3rd parties.
After reading through this thread, do you not think you might be being overly selective? You said you're only really applying to European jobs but hiring people that need working visas is usually a pretty big detractor. No LinkedIn can work, but in the LLM era you need something to prove you're real and you also stated you haven't been asking previous coworkers for referrals. It's very normal here to ask coworkers for referrals so I assume it's a cultural difference? But they usually get you to the interview stage. Your resume is likely not baring you but if you are applying to large enough companies that have thousands of applications on each role, they are using LLMs to filter them down. You should be running your resume against an "Act as a hiring manager" style prompt to see if you're getting filtered that way.
The article seems to only depict it being similar to the H20 in memory specs (and still a bit short). Regardless, Nvidia has their moat through cuda, not the hardware.
It's interesting to see so many people convinced it's related to their specific media they saw (all unique from each other). I think this is more indicative that the issue is just well known and this is a response to the issue at large rather than a specific instance.
This is fair, I don't actually want no protection for creatives, but the enforcement of a law is a large part of what draws the line of generally good or bad for me.
Or to write it less pessimistically, the models are trained to prime their own context window such that by the end of the chain they arrive at more valuable responses. By creating intermediary steps in the chain, the next step is easier to generate rather than moving directly to the desired response. We call it reasoning because it is intuitively analogous to human reasoning methods though it is understood that LLMs don't succeed as generally as humans are able to.
Realized I didn't address your concern, I have still seen many creatives able to profit without strong protections. The gaming industry often has controversy with its drm but even drm free releases are continuously able to make large sales volumes and drm has been shown in that EU study from 2015ish to not actually protect sales.
Personally I strongly detest ip law, but I see it as problematic for its implementation rather than its intent. So much so that I would rather it be abolished than hold its current form. I feel this from the perspective of a creator and a consumer because I so often see works that are unfairly removed or have their profits stolen as a result of the guilty until proven innocent style enforcement. This makes me fearful of publishing works that build upon previous work (which many creative fields naturally do). The most blatant abuse I've seen recently was the Gamer's Nexus documentary being taken down by Bloomberg. However, IP is arguably more problematic in industries like medicine where patents are renewed ad infinitum by evergreening to milk a protected monopoly.
I'm in the exact same position down to the Pixel 7 pro. It feels like my options are either to move to graphene os or move to iOS. But grapheneos seems like it'll cause headaches that I am not that interested in dealing with, however my wish to escape vertical integration which plagues tech ecosystems still tempts me.
This is probably the culmination of every administration after Clinton. Unsustainable spending that keeps going until now we are feeling the pressure. This administration is definitely acting in ways I feel are significantly suboptimal but it's unlikely to show the deeper impacts so quickly.
I would go further to say that anki just gets an update where you can now have more dynamic context surrounding the words on cards. Just bold the target word. The same font is also something that can be easily fixed.
Personal opinion, I'm bearish on the shovel seller long term because the companies that are training AI are likely to build their own hardware. Google already does this. Seems like a matter of time for the rest of the mag 7 to join. The rest of the buyers aren't growing enough to offset that loss imo.
The alternative was likely that these people would move to platforms that actively prey on this instead. Imo it's good that they get people onto the newer model and help them bridge the way to healthier conversations. Though realistically, these people should just be using the personality tools and the default should be robotic so new people don't join the problem.