proc0

20 ms·

241.
by proc0 2y ago
"low-carb" here is relative. The typical diet, and also the majority of items at the average supermarket are full of carbs. A "low carb" diet is probably a normal level of carb intake compared to the diets of our ancesto
242.
by proc0 2y ago
The conclusion is partially correct in my experience. At the end of the day is has to do with making more money and it just happens that more features allows big companies to keep selling more of it, or at least that's the claim. I dis
243.
by proc0 2y ago
Agreed, and to add to the second point, in order to build it we must first understand it, although there is a chance to stumble upon it, but LLMs have already shown they are nothing like biological brains. One example is how brains use spat
244.
by proc0 2y ago
An actual artificial intelligence would have a logical underpinning to its computation. Computers can execute any logic really well. Intelligence depends on logic. LLMs however, are not using logic to arrive at any conclusion, they are just
245.
by proc0 2y ago
I think in general it is not hyped, but the current wave of it is. Right now the cutting edge is LLMs, transformers and architectures around that. I think it's possible it will reach a limit in the next year or two, and even though tha
246.
by proc0 2y ago
I think WFH is another point of leverage for employees. If something doesn't work out with the job it's easier to deal with it on your own terms, whether that's taking more breaks or silent quitting. This leverage is exactly
247.
by proc0 2y ago
Can they not solve this problem? This shows the lack of creative problem solving that is widespread across companies. They've known only one way to work and they're convinced it's the only way, when in theory there is probabl
248.
by proc0 2y ago
Yes! For the same reasons engineers shouldn't test their own features. It's really an engineering 101 to understand that it's much safer and better to eliminate the inherent bias from testing your own solution. The reason com
249.
by proc0 2y ago
Where's the github/source link? Anyway this sounds really cool, but it's nothing that new for anyone that has done FP for a while. I've been trying to get companies to do something like this for a while but it's jus
250.
by proc0 2y ago
DEI is just a deeply flawed idea. If there isn't enough representation in some field, the solution is to go to the group that lacks representation and first convince them they should try and build a career in said field, and second, he
251.
by proc0 2y ago
Here's the crux of why AI is not yet useful for more than simple projects: they cannot actually know when something is correct or wrong. As a result there is no guarantee that something is implemented properly. To clarify, a human juni
252.
by proc0 2y ago
As people have pointed out it depends on the details. Maybe SSN is not the primary key and it's something else like name + DOB. It still requires more details to prove that it's being used maliciously but there's no reason to
253.
by proc0 2y ago
I can grant all of this makes sense for contractors, given that it's not building a career within a single company. I was more talking about full time, long term careers within a single domain. > No matter how good your code is, hig
254.
by proc0 2y ago
> Do you think “engineering” is only the physical labor? Do aircraft engineers and building engineers actually do the construction work? No, but software is inherently different because you can leverage existing software to create more s
255.
by proc0 2y ago
Right, but it's an example that illustrates a broader problem with how software still has the same issues it had for decades, especially in network programming and web. In contrast hardware has made huge advancements.
256.
by proc0 2y ago
Right, being senior and above basically means doing less engineering. I hope that's something we can agree on because otherwise we would need to discuss the semantics. > All of the things I listed I could do - project management, ba
257.
by proc0 2y ago
> That’s it, there is no other reason for a company to pay anyone except to bring more value to the company than they cost to employ them. The implicit part is that if you engineer a good system then it saves money with less bugs and bre
258.
by proc0 2y ago
To clarify, I don't mean having no concern for the business or customer, but rather having that as a priority and the primary way to determine contribution. Engineers should care about the engineering first, which implicitly has value
259.
by proc0 2y ago
Why are they hiring the candidate to begin with? This entire "justify your existence every quarter" is one of the bigger downsides of how the economy works, because it allows people to be exploited. it's the metaphorical writ
260.
by proc0 2y ago
Yeah and I would extend it to a better company as well.
261.
by proc0 2y ago
I think you're conflating two different things. On the one hand you have how a product is made, and on the other there is the demand for it which will affect how the product is made. In the case of software these two things are unfortu
262.
by proc0 2y ago
> Many engineers would be happy to polish the thing that exists, perhaps adding complexity the form of scalability, modularity or reusability that isn't actually needed, and in fact may make it harder to adapt to broader changes in
263.
by proc0 2y ago
Yes, that's part of the reward, but normally you get raises, bonuses, etc. based on yearly or quarterly reviews. If you are not entry level, and you just focus on programming while getting a good pay, and you are not seen as underperfo
264.
by proc0 2y ago
Yes with promotions, bonuses and overall good reviews, which are only marginally based on technical achievements (when it should be absolutely based on that).
265.
by proc0 2y ago
There is implicit impact by engineering good systems and software. It's like hiring a plumber and telling them "you need to improve my house" rather than "you need to improve the plumbing in my house". When your obj
266.
by proc0 2y ago
And rewarding engineers for "impact" rather than correct, reliable, and extensible software.
267.
by proc0 2y ago
I propose Solaris.
268.
by proc0 2y ago
I'm seeing a lot of postings for 5+ years of experience, so it doesn't seem that bad. That said I'm not applying unless it's a great fit, so I'm not yet aware of how much competition there is (I've just begun l
269.
by proc0 2y ago
I don't know that, mostly speculating based on how mixture of experts is outperforming decoder-only architectures, which means we're already composing transformers to squeeze the most out of it, and still it seems to fall short. T
270.
by proc0 2y ago
This is lowkey cope. AI should be like talking to another human, at least that is the promise. Instead we're getting glorified autocomplete with padded language to sound like a human. In its current form LLMs are pretty much at their l

More ›