sealeck

6 ms·

1.
by sealeck 1mo ago
Postgres has also had pretty serious bugs, eg fsyncgate
2.
by sealeck 2mo ago
Trains are literally scheduled according to a timetable... traffic isn’t
3.
by sealeck 2mo ago
It’s interesting to claim that the ‘tech bros’ oppose hardware attenuation and age verification when this will massively benefit them; everyone will be forced to use their operating system and the government will have exercised its power to
4.
by sealeck 2mo ago
There are few occupations in which a man can be more innocently employed than the pursuit of devising asynchronous network protocols?
5.
by sealeck 2mo ago
And we can also ignore model law; we should require OpenAI/Anthropic to provide unrestricted access (at standard API rates) to their competitors so they can use this to train new models.
6.
by sealeck 2mo ago
I think this comment is quite disingenous -- it's like if there's a rule that says "nobody can walk on the grass" that you object to because you'd like to have a picnic with some friends; your claim is that if someo
7.
by sealeck 3mo ago
> lack of the path taken from one point to another limits your ability to train models that can efficiently recreate the work Isn’t this the problem inference (training) a model is designed to solve :)))
8.
by sealeck 3mo ago
Have we reached the limits of scaling? Sadly it appears that larger model still equals better model
9.
by sealeck 4mo ago
As in cum-"captain of industry"
10.
by sealeck 4mo ago
I think the issue with these kind of stances is that they are basically status quo bias; why don't you object to the computer itself, and thus refuse to write programs? After all: they were invented by the UK military in the pursuit of
11.
by sealeck 4mo ago
Allegedly OpenAI's contracting model is much more vicious than Anthropic's; at work (admittedly a little IP-protective) we have unlimited Claude, but no Codex subscription because OpenAI won't give us sufficient guarantees ar
12.
by sealeck 4mo ago
> The technologists who create it believe they should control it I think there's an interesting phenomenon where it is _not_ the people who control it, but instead a kind of international finance man cum-captain of industry (perhaps
13.
by sealeck 5mo ago
Why does the false positive rate matter if you have a verifiable oracle? You can just disregard anything that fails the oracle
14.
by sealeck 5mo ago
Dystopian present
15.
by sealeck 6mo ago
Surely the other way around? Phone QA process >>> disposable vape QA process...
16.
by sealeck 6mo ago
Yes and making a horse drawn cart drive itself was thought to be impossible so why don't we have faster than light travel yet...
17.
by sealeck 6mo ago
> I found an (as of yet not-root-caused) error in sqlite (no crash or coredump, just returns the wrong data, and only when using sqlite in ram-only-mode). You should report this to the SQLite developers - they are very smart and very int
18.
by sealeck 7mo ago
> Stallman has always been right. It's mind boggling just how right he was about everything. Mind boggling right about not allowing GCC to be used as a library, his comments on Jeffrey Esptein, a refusal to in any way compromise (e.
19.
by sealeck 7mo ago
> Jane Street skims money from our retirement accounts by building expensive clocks that the rest of us don’t have access to and adversarial queue modeling How does Jane Street skim money from those who hold passive index funds?
20.
by sealeck 7mo ago
> The company needs a certain amount of productivity at each point. Um, no?
21.
by sealeck 7mo ago
> I reckon this move is related to bitcoin doing poorly. A LOT of their revenue is bitcoin related and I reckon they realized they're going to have an absolute stinker of a Q1 '26 result... I had to look this up - in the last 1
22.
by sealeck 7mo ago
> Musk has shown that Twitter can operate with 5% (approximately?) of the workforce he inherited Is X profitable? I don't think the argument was that Twitter couldn't _operate_ with 5% of the workforce (i.e. skeleton sysadmin c
23.
by sealeck 7mo ago
> instead of, well, producing more, better products faster, thus increasing its competitiveness? Probably because this is not Block's business strategy. If they could do this, then they would...
24.
by sealeck 7mo ago
> probably the most honest message you’ll ever see Interesting that this is your takeaway; it seems that this is effectively an investor-friendly way to admit that Block hired too many people over the course of the pandemic and doesn&#x2
25.
by sealeck 7mo ago
> there will be copyright issues Next.js is MIT-licensed. Cloudflare's rewrite is... also MIT licensed...
26.
by sealeck 7mo ago
There are many really excellent papers out there - the kind which will save you hours/months of work (or even make things that were previously inviable to build viable). That said, it is amazing how terrible a lot of papers are; people
27.
by sealeck 7mo ago
Why is this LLM spam?
28.
by sealeck 7mo ago
> Its like going 70 mph in a sleepy subdivision because a road sign on the interstate says you can go 70 there. > Trump is taking an law that says "You can do X if Y" and saying "I can do X" I think it's more
29.
by sealeck 7mo ago
Do you have examples of deadlocks/livelocks you've encountered using SERIALIZABLE? My understanding was that the transaction will fail on conflict (and should then be retried by the application - wrapping existing logic in a retry
30.
by sealeck 7mo ago
I think it's a bit hard to say that this is definitively true: people have always been interested in running linear algebra on computers. In the absence of NVIDIA some other company would likely have found a different industry and sold

More ›