Topfi

12 ms·

211.
by Topfi 4mo ago
I just checked and it's 54% in the US [0] vs 0% in France (cause basic public healthcare isn't tied to employment). So, unless you are referring to absolute numbers (not very helpful when comparing countries of different sizes), I
212.
by Topfi 4mo ago
I doubt that the capabilities of GPT-5.5-cyber aren’t known by the US government considering OpenAI is their primary LLM partner after Anthropic had concerns about using models for autonomous weaponry and mass surveillance of US citizens. I
213.
by Topfi 4mo ago
I can, just not in the US [0]. I always presumed this is linked to the health care being provided by employers rather than having a more robust safety net that allows for civil disobedience without having to fear existential risks. However,
214.
by Topfi 4mo ago
Anthropic is one of the two consistent revenue sources for XAI via their colossus deal. I have been critical of this man longer than most, but I don’t see him hurting his own bottom line.
215.
by Topfi 4mo ago
SSRIs are a first line treatment across many EU countries too, yet we somehow manage.
216.
by Topfi 4mo ago
I still am struggling to understand why they informed the government about something that is known to be an issue in every LLM. There is no LLM that cannot be jailbroken, so unless this means that we have reached the absolute maximum public
217.
by Topfi 4mo ago
Great article, the worst offender is compact tab mode in the current Safari. The animations they implemented make that unusable, sometimes it’ll move tabs away from where the tab was when clicking, the animation always look clunky and the e
218.
by Topfi 4mo ago
By how, I meant specifically the internal activations, which no person in the field claims to have a comprehensive understanding of, not next token prediction as the underlying technology. The whole interpretability of it all is the crux I
219.
by Topfi 4mo ago
My personal issue in comparing LLM progress and risk as labs publicly predict it with nuclear power in the middle of the 20th century is that the processes by which it works where fairly quickly well understood and the risk could thus be re
220.
Can $100 ChatGPT and Claude Fable Solve PhD Math? [video](youtube.com)
1 points by Topfi 4mo ago|0 comments
221.
Riemann-Bench(surgehq.ai)
2 points by Topfi 4mo ago|0 comments
222.
Claude Fable 5: the first public Mythos-class model(artificialanalysis.ai)
4 points by Topfi 4mo ago|0 comments
223.
$250M iPhone Settlement Proposed in Apple Lawsuit over Misrepresented AI(classaction.org)
2 points by Topfi 4mo ago|0 comments
224.
Apple Game Porting Toolkit 4(developer.apple.com)
2 points by Topfi 4mo ago|0 comments
225.
by Topfi 4mo ago
> Should you be able to use a Samsung SoC in an Apple phone? How did we go in less than two comments from providing access to APIs that are already present, implemented and actively used by Apple (who in their holy wisdom deem us mortals
226.
by Topfi 4mo ago
Great effort and a bit closer to my private evals than DeepSWE. I greatly appreciate the focus on false negative and positives, along with simply being far more focused on actual, mergeable quality output over plain passing. Could see a lot
227.
by Topfi 4mo ago
> […] fed large chunks of an article multiple times to an LLM […] So they had to prompt? An LLM? I got this argument before and still don’t get what it’s trying to say. These models do not output anything unless prompted, that’s not any
228.
You can now use your Gmail account in Proton Mail(proton.me)
5 points by Topfi 4mo ago|0 comments
229.
by Topfi 4mo ago
50+TFlops is nothing, I got that in my MacBook, but besides that, when, a few years/decades from now, whatever arbitrary compute limit we think prevents Armageddon comes down to enthusiast and consumer level, what then? This isn’t Uran
230.
A company didn't spend $500M on Claude in a month(stax.sh)
2 points by Topfi 4mo ago|1 comments
231.
by Topfi 4mo ago
Good. I firmly am of the opinion that if any ADAS sold to require limited intervention (Supervised FSD, Mobileye, Nvidias offerings, etc.) actually was reliable and as safe or safer as a human (even if only the former, there is no fatigue,
232.
by Topfi 4mo ago
The fact that Anthropic models are offered at the same API pricing by not just themselves but AWS, Azure and Vertex despite Anthropic taking a major slice on licensing along with the cost an open weight 1T parameter model like K2.6 costs to
233.
Casting for Foot Orthotics: A Review of the Clinical and Cost-Effectiveness [pdf](cda-amc.ca)
2 points by Topfi 4mo ago|0 comments
234.
Firefox AI Coding Policy(firefox-source-docs.mozilla.org)
2 points by Topfi 4mo ago|1 comments
235.
by Topfi 4mo ago
I'd love to be a fly on the wall at some dashboard SaaS companies today. Especially with AWS for OpenAI models on top, the moat that I've been told by quite a few friends in the sector that was deployment is likely busted, as was
236.
Nvidia's New Old Chip – Nvidia RTX Spark Analysis [video](youtube.com)
3 points by Topfi 4mo ago|1 comments
237.
by Topfi 4mo ago
Title was editorialised slightly by adding "Nvidia RTX Spark Analysis". Youtube titles are unfortunately rarely as clear.
238.
Interop 2026: Continuing to improve the web for developers(web.dev)
2 points by Topfi 4mo ago|0 comments
239.
by Topfi 4mo ago
>> The EU is even building their own Starlink equivalent. > No they aren’t. That’s exactly what IRIS [0] is though… [0] https://defence-industry-space.ec.europa.eu/eu-space/iris2-s...
240.
by Topfi 4mo ago
If I may slightly tweak your example to highlight why I find it very flawed: > It's April, 1891. Magically, a drone swarm with lights piloted to show a face [0] materialises above London. Hidden speakers command the public to listen

More ›