dan_mctree

8 ms·

31.
by dan_mctree 3y ago
For a lot of companies, switching databases is effectively impossible. Perhaps not quite as impossible as switching game engines, but certainly impossible enough that it would kill the business. A lot of companies use database specific func
32.
by dan_mctree 3y ago
I don't know why but no one hires or listens to the architect who recommends sticking to safe monolithic-esque systems. If you're not talking cloud and microservices and the newest unproven frameworks, you're an old fogey who
33.
by dan_mctree 3y ago
There's not nearly enough transformers in reserve to properly rebuild the grid and you can't always just drop in and replace either. New transformers are also not currently able to be produced fast, and having society collapse aro
34.
by dan_mctree 3y ago
Interesting idea, of course it seems very unlikely, but running with it, what could potential tells be? I imagine there might be tells something weird is off if you got really close to them, so they may stay on the periphery of most humans
35.
by dan_mctree 3y ago
Hey this fits my entirely unsubstantiated pet theory: evolution is not just partially Lamarckist, but there's also a purposeful mental component to it. Animals can -presumably subconsciously- partially pass on traits that they themselv
36.
by dan_mctree 3y ago
Cooling my forehead by applying water has helped me fall asleep, especially on warm days. I suspect an inability to fall asleep plus chaotic mind wandering is often related to an overheated brain and that this is a good artificial way to he
37.
by dan_mctree 3y ago
Can we all just be honest and just agree that prohibitions do in fact make access to the prohibited thing harder and reduce the prevalence of it. This is true whether your topic of choice is drugs, guns, abortions, alcohol, some form of sex
38.
by dan_mctree 3y ago
If you only offer creditcard payments and want international customers, you will missing out on a lot of customers. For example in the Netherlands, iDeal is the standard of payment and many Dutch customers do not have other payment options
39.
by dan_mctree 3y ago
Maybe they could attach a cable and drag it up to a ship? I'm sure there's ways to help these people with joined effort
40.
by dan_mctree 3y ago
And I'm done with developers who proudly boast of 95% test coverage using mocked tests because "we don't want to test the database" and then have database changes cause production crashes. "But my tests were all gre
41.
by dan_mctree 3y ago
We do know that AI's have at least some advantages. For example they're more accurate and orders of magnitude faster at floating point calculations. They'll also have communication interfaces with other (sub)systems that are
42.
by dan_mctree 3y ago
Not as extreme, but garden path sentences sometimes give me a glitchy feeling in how they lure you into an parsing error that takes difficulty to overcome. For example: The complex houses married and single soldiers and their families.
43.
by dan_mctree 3y ago
It's possible that the crafts are not designed for the earth environment and that it is different enough from expected environments that there's some ways for these to crash. For example, they may not be able to handle earth storm
44.
by dan_mctree 3y ago
Imo the most likely scenario if it's true, is that these are automated von Neumann probes that traveled here without faster than light propulsion and that they're here merely to explore, observe or possibly self-replicate. I belie
45.
by dan_mctree 3y ago
They could be von Neumann probes having travelled for a long time. Even without FTL travel, von Neumann probes could colonize an entire galaxy in reasonable time (a few million years). Even just a single successful civilization somewhere in
46.
by dan_mctree 3y ago
Imo its this type of self replicating AI that is likely to cause the first global AI catastrophe. A worm-like AI, even if not close to human level intelligence, could if autonomous and sufficiently good enough at hacking, infiltrate a huge
47.
by dan_mctree 3y ago
What exactly is the ideal sort of hardware to be able to run and train large models? Do you basically just need a high end version of basically everything?
48.
by dan_mctree 3y ago
Someone ought to start a charity to hire personal assistants for people like Terence Tao. Or get some government tech grant working on that. That would be probably one of the most efficient investments in progress you can get
49.
by dan_mctree 4y ago
The bad guy with an AI may very well build such a competent and fast acting AI that there's no defense possible. To strain the analogy, if the good guy with the gun already has a bullet in him, he ain't stopping much
50.
by dan_mctree 4y ago
This is how a lot of the younger adults are keeping their head above water. Not everyone has this luxury though, it's not like everyone's parents made bank
51.
by dan_mctree 4y ago
>I ask, how. A superintelligent AGI could easily follow this three step plan: 1. Optional: Overtake and spread computation to security vulnerable computers (presumably, basically every computer) 2. Gain a physical presence by convincing
52.
by dan_mctree 4y ago
Never quite got this argument. You can frequently do 1000 fold improvements or do even better by optimizing within the same algorithmic complexity. That frequently flips the bit of "is this software usable". Sometimes you can even
53.
by dan_mctree 4y ago
>It's likely the planet will become hostile to (human) life long before AI will be able to do us any harm. Sure, because the planet is already hostile. However, AI could potentially do much more harm much more quickly if we're
54.
by dan_mctree 4y ago
At this point ChatGPT is blowing our human ability to use language so far out of the water on so many levels, I'd argue we should start putting quotes around our human ability to "understand" language. GPT-4 has already eclip
55.
by dan_mctree 4y ago
How about just picking a non-cyclical time? Just count the amount of seconds/milliseconds since 1 Jan 1970 where the duration of the second is synced to those on earth. Just have no official days/hours/years whatever, no leap
56.
by dan_mctree 4y ago
As someone who mostly just runs into SO through searches and rarely asks: I want to see these duplicates thrive, I want to see the ideas. It's so common that the first SO thread is useless but that a near duplicate does actually provid
57.
by dan_mctree 4y ago
It's easy to underestimate just how difficult it is to get your context indexed and ranked well. Personally I was under the delusion that if you make a solid website, you'll get some organic visitors through search engines from pe
58.
by dan_mctree 4y ago
For me I don't believe there's a way to make the activation small enough of a thing that I would be able to consistently do it. If my task were to consistently lift a finger every Monday I don't think I'd be able to do i
59.
by dan_mctree 4y ago
ChatGPT can provide correct citations because somewhere deep in its weights it does lossily encode real texts and citations to real texts. That makes real citations in some cases be its most confident guess for what is supposed to come next
60.
by dan_mctree 4y ago
This is absolutely stunning. I'm beyond curious what is possible in this space. Could we potentially start regenerating and even redesigning lifeforms, including humans? How about inducing growth of a second brain? Or perhaps we could

More ›