That seems like a pretty straightforward issue to solve, to simply backup only those files that are actually on the system, not the stubs. If it's on your computer, it should able to get backed up. If it's just a shadow, a pointer, it doesn't.
Making the change without making it clear though, that's just awful. A clear recipe for catastrophic loss & drip drip drip of news in the vein of "How Backblaze Lost my Stuff"
Not what I meant: The other cloud storage services connected to the computer, eg OneDrive. Those files, when they are just stubs. I'm saying that Backblaze could simply not backup stubs, if the person isn't syncing the actual file to their drive. If they are, backblaze should back it up.
The point of the stubs is so you don’t have to know which cloud files are actually on your device at any given moment, because they will be fetched automatically. Marking files as “always local” is a niche feature, and in any case has nothing to do with whether you want those files backed up.
To have this thing you’re not supposed to need to worry about affect whether your files got backed up is exactly the problem here. The goal is to back up your files, whether they’re in the cloud or not.
I sympathize with Backblaze’s problem with their file change monitor, but then they should considee implementing connectors for OneDrive, Dropbox, etc. and back up files directly from the cloud.
“We only back up stuff from your computer” has always been their stance, and clearly it’s a way to reduce their costs. I can understand not wanting to engage in the “please backup these 10TB I have in Dropbox” requests.
I think backing up the materialized files is appropriate. That’s what they (used to) promise.
Imagine if they could detect stab or real file huh? Space technology, I know! Or just fucking copy them as stubs and what's actually downloaded as actually downloaded! Boggles the mind!
Or maybe just do what they do now, but WARN about that in HUGE RED LETTERS, in the website and the app, instead of burying it in an update note like weasels!
No, no-- I've seen the movie and I'm pretty sure it was established that the only winning move was not to play. Not ruling out the possibility I'm misremembering, there was more than one game in the movie, it could have been Galaga?
Tic-Tac-Toe is the existing game the machine realises can't be won, and then Global Thermonuclear War is the next game it simulates and discovers also cannot be won according to the metrics it is using.
Hi Tim, I'm not sure what role would suit me best- titles haven't fit me well over a 15+ year career, but "data science and adjacent" wouldn't be inaccurate. With a focus on translating technical findings into strategic input for leadership decision & policy making- neutral or opinionated per the context's needs. The former has usually intersected with operations in coordination with IT, the later to the c-level.
By training: dusty philosophy & cog sci, less dusty linguistics, translation theory, NLP. More recent work w/ London School of Economics on AI Law, Policy, & Governance. LLMs come naturally, so I'm looking for a change. (Location: NYC area, USA)
For what I can offer:
Over the past few months I've built a toolkit for model interpretability that operates below black-box prompt-output observation and raw numeric activations, collapsing a lot of complexity into something more discrete and tractable, with accuracy that appears, at least from outside the core AI industry, to be harder to come by with many current methods. This includes monitoring and inference-time intervention without retraining or weight modification.
I've used it to improve benchmark performance across modalities- DeBERTa mini boosted overall > 10% in the adversarial HANS dataset through fewer false positives, no retraining or degradation in other performance. Needs testing in any deployment of course. Similarly, MedGemma, MMFLD, Whisper, a few others, with some of their standard benchmarks. Same methodology of exploration and inspection.
Utility tooling along the way includes an intuitive REPL interface for token-by-token exploration of model internals during inference, or optionally post-inference with data capture by SQLlite & LanceDB, analysis with UDF's and python. Other tools for pre-token-gen semantic monitoring and intervention intra inference. I've observed some things that are more speculative, though still promising, for understanding model behavior. All generally grounded in classical Linguistics areas of study that seem less mined for insights than industry has had opportunity to pursue in-depth with but actionable nonetheless.
I'd love to talk further. These are mainly my personal time interests though, so a resume through a job posting doesn't generally cover things adequately, if a different option is available, though I can go that route if you'd prefer.
That linked article says its about RLVR but then goes on to conflate other RL with it, and doesn't address much in the way of the core thinking that was in the paper they were partially responding to that had been published a month earlier[0] which laid out findings and theory reasonably well, including work that runs counter to the main criticism in the article you cited, ie, performance at or above base models only being observed with low K examples.
That said, reachability and novel strategies are somewhat overlapping areas of consideration, and I don't see many ways in which RL in general, as mainly practiced, improves upon models' reachability. And even when it isn't clipping weights it's just too much of a black box approach.
But none of this takes away from the question of raw model capability on novel strategies, only such with respect to RL.
Pithy. But a made up quote by Tytler, he never said or wrote that.
Tyler expressed some skepticism of Democracies but nothing like this. The too on-the-nose nature of this often passed along bit of propaganda should also be the giveaway that it might be one of those rare things on the internet that someone may have been less than honest about the origins, and go look and see.
so… choices, as you see them in this issue, the lenses through which on the one hand you think is extreme and the other appropriate… are either screens-as-drugs or sports fishing?
Some middle ground might be there somewhere. But if forced to choose… the choices for interpreting behavioral engineering funded by $billions in research for over a decade + data harvesting on a scale unprecedented, for the purpose of manipulating users:
There were no uncomfortable truths there about code agents, save one of the 4 points which was that maybe they sometimes get prompt injected if you let them search for things online and don't pay attention to where they search and the code they write. That's not an uncomfortable truth in the normal sense of "I know you don't want to admit this but..." and more just the thing that, if you didn't know it already 8 months ago, you certainly should by now.
The other truths that were not about coding agents:
--Skill Atrophy. (Use it or lose it-- another thing we already know)
--The economics of serving code agents at scale (Ungrounded in actual numbers, only OpenAI's miscellaneous statements and annecdotes. Actual cost of running code agents: last gen's mid-tier gaming gpu's will get you reasonably close to Claude Sonnet if you put just a little time in to an agent harness, and its getting cheaper and cheaper for better and better. So, at scale, with real sysadmins doing the hard engineering to eek out every last bit of performance-- well, infra needed for serving these isn't the cost center)
--Copyright. (This passed on the same bad read of a court ruling half the press has been doing for a few years now. TLDR: The Thaler vs. Perlmutter case, which said nothing about output not being protected by copyright. It denied Thaler's attempt to register *the AI* as the owner of the copyright)
Good luck getting the average person through the setup process
AI is part of the problem with what MS has shoved in to things but it may be part of what can help with the underlying issue of this behavior by corporations.
The average user increasingly will not need to be walked through in certain ways, they’ll only have to be aware something, some way, is possible. Because we are most of usthe average, meaning outsider to knowledge and understanding of things their functioning on a computer. I can strip out tired windows behavior to some extent and certainly stand up a Linux desktop. But I didn’t know how to easily manage retrieval of data from an old disc image that refused to mount. But I knew it was there and not impossible so I asked Claude. A one shot prompt that a few minutes later had Claude reading raw bytes in someway and finding the location of a few files I needed.
So there is potential for AI to fill some gaps in this way and make some things easier and more in reach of average users. It’s potential only though, so continuing to work and ensure open models remain a thing, it’s important. Just like the Internet enabled a lot of things previously out of reach of people. And yeah, that was not an un mixed blessing with the rest, so all the more reason to move forward thoughtfully.
Making the change without making it clear though, that's just awful. A clear recipe for catastrophic loss & drip drip drip of news in the vein of "How Backblaze Lost my Stuff"
reply