Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
trod1234
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
72 ms
·
691.
▲
by
trod1234
2y ago
You are mistaken. Dig into tracker-miner-fs, d-bus activation re-parenting, as well as the snap callbacks that can trigger arbitrary local code execution. Usually found in a directory under /usr/libexec in addition to its regular
692.
▲
by
trod1234
2y ago
Not likely at all. Person A. B is controlled by A, and C impacts A's control of B. The concept of B solving C, while at the same time having no impact on C, is patently absurd (contradictory), and the reasoning path demonstrably circu
693.
▲
by
trod1234
2y ago
Most likely, starting in 2025, we will have hit a point of no return. It won't be considered this until much later, if there are people who manage, but still it will be a great year. Great... but terrible things will be coming. AI will
694.
▲
by
trod1234
2y ago
I think you are overthinking this. Consumers rely on advertised claims being truthful. It is not a matter of people badly predicting their own needs in most cases,though there are some that do have problems with this. It is a matter of peop
695.
▲
by
trod1234
2y ago
That is categorically untrue. The internet was available well before 94. AOL and Compuserve were available and provided access to quite a large number of people starting in 85 iirc, and by 89/90 their free month promotions targeting fa
696.
▲
by
trod1234
2y ago
This is a recipe for disaster wherever it might occur, no good can come from it. https://www.washingtonpost.com/technology/2024/12/10/charact... LLMs are not people, only people know how to give guidance
697.
▲
by
trod1234
2y ago
Many would disagree, solely on the basis of the many knock-on effects where such companies in other areas have cut corners with egregious failures. > Augmenting a parent so their child has a world class tutor is incredible stuff. So does
698.
▲
by
trod1234
2y ago
I have no children yet, but I would never consider introducing LLMs until my children reach a point of maturity where they can accurately detect sophisticated deception. Even after that, I would consider these constructs as dubious in natur
699.
▲
by
trod1234
2y ago
I hate to reign on your parade but Canonical (Ubuntu) is doing the same thing as Microsoft just to a slightly lesser extent. The same pretty much goes for any of its many derivatives. Redhat is no longer open. Arch and alpine based have har
700.
▲
by
trod1234
2y ago
Lawmakers have created endless nests of ordinance, law, etc deferring responsibility to other aspects of government. Regime change plans generally involve sabotaging resilient systems to make them become brittle. Then you force those system
701.
▲
by
trod1234
2y ago
I suppose that would depend on your working definition of correctness. If correct means its legally unenforceable and practical and you only take that into consideration in isolation, then yes, right up until logarithmic error occurs, as yo
702.
▲
by
trod1234
2y ago
There is no real future in AI long term. Its use caustically destroys more than it creates. It is worthy successor of Pandora's box.
703.
▲
by
trod1234
2y ago
This opinion article lacks appropriate backing for the narrative pushed to be compelling, the data they reference is not included or linked, and the article suggests that non-enforcement of antitrust is the driver for food deserts which sta
704.
▲
by
trod1234
2y ago
> As soon as the general public cares and forces the issues. That is not how oligopoly and antitrust work. If producers don't actually make products that have the features available, there is no actual market for those products, and
705.
▲
by
trod1234
2y ago
While an interesting perspective, this HN article is a poster child for how not to resurrect old blog posts. Initially, I was not sure how much credibility to give this because there were no corroborating records for the time period specifi
706.
▲
by
trod1234
2y ago
Chain of Trust is the low hanging fruit, there are many other potential avenues that compromise TLS. If you want to see a full discussion of this exact topic by Cybersecurity professionals, a reddit post covered it a few years ago. I'l
707.
▲
by
trod1234
2y ago
In my opinion, your setup is likely to be insufficient for the purposes you want, and in some minor or not so minor ways may be more likely to draw additional scrutiny (i.e. grapheneOS (minor)/tailscale(?)). Physical access is almost n
708.
▲
by
trod1234
2y ago
A decent article, although it doesn't touch on why or how the reader who wrote in developed those feelings of guilt, which is the more important issue. The article covers the cursory salient points within philosophy relating to Kant, a
709.
▲
by
trod1234
2y ago
> Since about the 1960s, most of the damaging spies have been in it for the money, not ideology. That isn't true as a general statement. It is true that American's in particular are generally more open to the money than other m
710.
▲
by
trod1234
2y ago
That line of reasoning neglects the fact that there is very little control in such systems that differentiate between the subspecies. CRISPR for example has been hailed as surgical tool for slicing DNA, and works well in controlled environm
711.
▲
by
trod1234
2y ago
The underlying content is largely irrelevant. What is relevant is what is advertised, and for what purpose. In many cases this alone may be sufficient to show intent/malice/negligence, especially when it breaks law. I don't i
712.
▲
by
trod1234
2y ago
There is no reassurance that can be made since there have already been examples of voice cloning that are not differentiable from the real people. There is a very common saying in IT, "Attacks only get better". Voice Cloning falls
713.
▲
by
trod1234
2y ago
> not worried ... but the creator of this website should be? They should be worried. Hammers have many legitimate uses, voice cloning not so much. Claiming its a tool is meaningless when it has few legitimately legal uses. If there is an
714.
▲
by
trod1234
2y ago
This wouldn't have an impact. There would inevitably be a classifier that acts as a filter pre-training to identify slop and ignore it.
715.
▲
by
trod1234
2y ago
I happen to fall into the same group as the previous poster who is a voice actor. I know you did this as a challenge for fun but in fairness you should keep those challenges to yourself and not publicize them when the technologies involved
716.
▲
by
trod1234
2y ago
I don't know about their business suite. I've used their calendar and email. The calendar has had issues over the past few months where it would revert unexpectedly and silently back to a previous date (like they reverted to backu
717.
▲
by
trod1234
2y ago
This is why they chose to put AI into word processors and excel. So they can take the PII in a derived but reconstructable form (weights).
718.
▲
by
trod1234
2y ago
Worse than exhausting, this is clearly a pattern of abuse done by purposeful intent. Security fatigue is a well known thing in IT. Configuration fatigue where your configurations malevolently switch back on after the options you chose, disa
719.
▲
by
trod1234
2y ago
ISPs would be required to disconnect people who are found to be connected to various torrent sites under the claim that they pirate, then it expands to perfectly legitimate sites, and then the internet becomes useless, and violence breaks o
720.
▲
by
trod1234
2y ago
I agree though not for the same reasons. Copyright law lacks the authority to deprive people of their constitutional rights. Those rights would inevitably be violated as many bureaucratic functions of government require internet access, and
More ›