Topfi

40 ms·

571.
by Topfi 2y ago
Counterpoint, ChatGPT as a brand has insane mindshare and buy in. It is synonymous with LLMs/“aI” in the mind of many and has broken through like few brands before it. That ain’t nothing. Counter-Counterpoint, I still feel investors pr
572.
by Topfi 2y ago
In fairness, the AGI definition predicted by doomsday safety experts who don’t have time for such unimportant concerns as copyright or misinformation via this tech, everyone who is most certainly not merely hyping to get investor cash and t
573.
by Topfi 2y ago
I tentatively agree that LLMs have reached somewhat of a ceiling at this stage and diversifying would make sense at this stage, in any other industry. But as others pointed out, OAI and others have attached their valuation directly to their
574.
by Topfi 2y ago
> There are no Alps between Vienna and Graz [...] You what? Please don't take this the wrong way, but the things I read about European countries on HN sometimes are asinine and part of why, despite desperately wanting to sometimes,
575.
Meta cheats on Llama 4 benchmark(heise.de)
3 points by Topfi 2y ago|0 comments
576.
Perceiver IO: A General Architecture for Structured Inputs and Outputs (2021)(arxiv.org)
2 points by Topfi 2y ago|0 comments
577.
by Topfi 2y ago
> … that issue. Maybe I am missing something, but what issue were they alluding to? Yes, databases (anything digital for that matter) are easier to transfer over large physical distances. Why would there be a need to ship or transfer any
578.
by Topfi 2y ago
Thanks for sharing. To me, purely on personal preference, the Gemini models did best on this task, which also fits with my personal experience using Googles models to summarize extensive, highly specialized text. Geminis 2.0 models do espec
579.
by Topfi 2y ago
To put that in context, Claude 3.5 Sonnet (new), a model we have had for months now and which from all accounts seems to have been cheaper to train and is cheaper to use, is still ahead of GPT-4.5 at 36.1% vs 32.6% in SWE-Lancer Diamond [0]
580.
by Topfi 2y ago
Not OP, but likely because that was the only metric/benchmark/however you want to call it OpenAI showcased in the stream and on the blog to highlight the improvement between 4o and 4.5. To say that this is not really a good metric
581.
by Topfi 2y ago
Considering both this blog post and the livestream demos, I am underwhelmed. Having just finished the stream, I had a real "was that all" moment, which on one hand shows how spoiled I've gotten by new models impressing me, bu
582.
Milei's Memecoin Scandal Explained [video](youtube.com)
3 points by Topfi 2y ago|0 comments
583.
by Topfi 2y ago
Being an accessibility solution for people with MND or similar afflictions is one of the few exclusively positive use cases "AI voice clone" technology has, as one's personal voice can be very strongly linked to one's pe
584.
Deep Fake Detector Extension by Mozilla Firefox(addons.mozilla.org)
67 points by Topfi 2y ago|36 comments
585.
by Topfi 2y ago
We have heard this before when 100k and 200k were first being normalized by Anthropic way back when and I tend to be skeptical in general when it comes to such predictions, but in this case, I have to agree. Having used the previews for the
586.
Voidzilla – I got scammed by Enron [video](youtube.com)
1 points by Topfi 2y ago|0 comments
587.
by Topfi 2y ago
Video discusses this discussion on Bluesky: https://bsky.app/profile/kathryntewson.bsky.social/post/3lgj... Title was slightly shortened from original (Microsoft Copilot Manager roasted, doesn't understa
588.
Copilot Manager doesn't understand Confidentiality [video](youtube.com)
2 points by Topfi 2y ago|1 comments
589.
by Topfi 2y ago
Honestly the first time I ever bought a game right on release day, cause I knew of the quality and wanted to support both GoG and hopefully in a small way signal to the industry that I want more of this in gaming.
590.
by Topfi 2y ago
Can also wholeheartedly endorse Alpha Protocol, also by Obsidian. A must play for anyone who feels RPGs are more than a few skill trees and binary choices, the way AP manages to interweave a massive complex conversations tree with the way p
591.
by Topfi 2y ago
Have yet to see a case where they didn’t also include the UK and other decidedly none EU states, which makes me think this has little to do with regulation and more trying to manage load during the initial release. Regulation is of course a
592.
by Topfi 2y ago
I see the most direct competitor in the Mac Studio, though of course we will have to wait for reviews to gauge how fair that comparison is. The Studio does have a fairly large niche as a solid workstation, though, so I could see this being
593.
by Topfi 2y ago
I'd love to own one, but doubt this will go beyond a very specific niche. Despite there being advantages, very few still operate their own Plex server over subscriptions to streaming services, and on the local front, I feel that the pr
594.
by Topfi 2y ago
> Legend goes that security oriented people will visually customize their machines with stickers (and their associated aging patina) and all kinds of digital cues on the different screens just to recognize if anything was changed. Maybe
595.
by Topfi 2y ago
I wouldn’t be surprised if similar will be found concerning the ARC challenge and it is why I still maintain my own private LLM challenges to gauge current capabilities. Course, I have little illusion that these are fully private, but it is
596.
Jsonevil – JSON non free license(wiki.debian.org)
2 points by Topfi 2y ago|0 comments
597.
Wayland and X – Daniel Stone (linux.conf.au 2013) [video](youtube.com)
3 points by Topfi 2y ago|0 comments
598.
Gell-Mann Amnesia Effect(en.wikipedia.org)
1 points by Topfi 2y ago|0 comments
599.
by Topfi 2y ago
Correct me if I am mistaken, but that seems to cover an IR based solution, which is not what was covered by OP or suggested in the Egg/Tweet that lead to the videos creation.
600.
MSN Premium Subscription(microsoft.com)
2 points by Topfi 2y ago|0 comments

More ›