thesz

10 ms·

151.
by thesz 8mo ago
> meaningless. No. You present "programs are graphs" as trivial truth. True trivial truths are, as you pointed out, meaningless. But you leave out degree of applicability - information in the dependence graph differs be
152.
by thesz 8mo ago
It does. Just like any other programming language.
153.
by thesz 8mo ago
Pipeline is a specific kind of a graph. Yes, there will always be hard-to-express queries, the question is how far can we go?
154.
by thesz 8mo ago
SQL is not a pipeline, it is a graph. Imagine three joins of three queries A,B and C, where first join J1 joins A and B, second join J2 joins A and C and third join J3 joins J1 and J2. Note that I said "queries," not "tables&
155.
by thesz 8mo ago
I see you and I raise approximate logic synthesis [1] [2]. [1] https://www.sciencedirect.com/science/article/pii/S138376212... [2] https://arxiv.org/abs/2506.22772 You can synthesize a l
156.
by thesz 8mo ago
Thanks! I haven't had my coffee yet. ;)
157.
by thesz 8mo ago
800 mm2, about 90mm per side, if imagined as a square. Also, 250 W of power consumption. The form factor should be anything but thumbdrive.
158.
by thesz 8mo ago
8B coefficients are packed into 53B transistors, 6.5 transistors per coefficient. Two-inputs NAND gate takes 4 transistors and register takes about the same. One coefficient gets processed (multiplied by and result added to a sum) with less
159.
by thesz 8mo ago
> what's your alternative? Brain derived neurotrophic factor (BDNF) serum levels are inversely associated with depression's severity [1]. https://pmc.ncbi.nlm.nih.gov/articles/PMC3188695/ Give
160.
by thesz 8mo ago
> Assume FP64 units are ~2-4x bigger. This is wrong assumption. FP64 usually uses the same circuitry as two FP32, adding not that much ((de)normalization, mostly). From the top of my head, overhead is around 10% or so. > W
161.
by thesz 8mo ago
Then you get to definitions like ": open ( string -- handle 1 | 0) ... ;" which describes returning algebraic type Maybe Handle unboxed on the stack. Algebraic types are fun, they can easily represent Peano arithmetic and get us
162.
by thesz 8mo ago
I cannot because these reviews are comments in Github PRs. I have to read them.
163.
by thesz 8mo ago
I had no expectations at all, I just asked questions, expecting answers. At the very beginning the tone of your comment, as I read it, was "agentic coding is nothing but simple, look they vote." Now answers to simple but important
164.
by thesz 8mo ago
You conveniently decided not to answer my question about quality of the solutions to vote on (ranking FizzBuzz memorization). To me, our discussion shows that what you presented as a simple thing is not simple at all, even voting is complex
165.
by thesz 8mo ago
> Then spin up a new agent to decide which approach of the three is best on the merits. Repeat this analysis in fresh contexts and sample until there is clear consensus on one. If there are several agents doing analysis of soluti
166.
by thesz 8mo ago
How several wrong assumptions make it right with increasing trials?
167.
by thesz 8mo ago
> It will accelerate the work and change your role from what it is today to something different; We yet to see if different is good. My short experience with LLM reviewing my code is that LLM's output is overly explanatory
168.
by thesz 8mo ago
> You should have test coverage, type checking, and integration tests that catch the edge cases automatically. You should assume that if you are going to cover edge cases your tests will be tens to hundredths times as big as the
169.
by thesz 8mo ago
> This capability has been demonstrated multiple times, specially when it was politically convenient, like for example the intercepted Hamas calls that showed that some of the rockets fall inside Gaza by mistake. Can it be someth
170.
by thesz 8mo ago
I took a look. Their system allow for intermediate state with errors. If that erroneous state can be stored to disk, they using a storage representation that is equivalent to text. If erroneous state cannot be stored, this makes Dion system
171.
by thesz 8mo ago
> Dion language demo (experimental project which stores program as AST). Michael Franz [1] invented slim binaries [2] for the Oberon System. Slim binaries were program (or module) ASTs compressed with the some kind of LZ-family a
172.
by thesz 8mo ago
Arithmetic coding of a single bit preserves ordering of encoded bits, if CDF(1) > CDF(0). If byte's encoding process is going from higher bits to lower bits, arithmetic coding (even with dynamic model) will preserve ordering of indi
173.
by thesz 8mo ago
Alphabet Inc, as Youtube owner, faces a class action lawsuit [1] which alleges that platform enables bad behavior and promotes behavior leading to mental health problems. [1] https://www.motleyrice.com/social-media-lawsuits&
174.
by thesz 8mo ago
Did he do so? I read his comment as a sad take on the situation when one realizes that one is talking to a machine instead of (directly) to another person. In my opinion, to participate in discussion through LLM is a sign of excessive LLM u
175.
by thesz 8mo ago
So I went to check whether LLM addiction is a thing, because that's was a pole around which the grandparent's comment revolves. It appears that LLM addiction is real and it is in same room as we are: https://www.mdpi.co
176.
by thesz 8mo ago
Fortran internally transforms expressions to assembly using symbolic computation. a+0->a, a+b->b+a. LLMs are very much supported by symbolic computation. These "arbitrary glyphs" are computed by symbolic computation. The out
177.
by thesz 9mo ago
Fortran is all about symbolic programming. There is no probability in the internal workings of Fortran compiler. Almost any person can learn rules and count on them. LLMs are all about probabilistic programming. While they are harnessed by
178.
by thesz 9mo ago
You are right about "LLM improving themselves" is impossible. In fact, LLM make themselves worse, it is called knowledge collapse [1]. [1] https://arxiv.org/abs/2404.03502
179.
by thesz 9mo ago
These diminishing returns are exponentially diminishing returns [1]. [1] https://arxiv.org/abs/2404.04125 " We consistently find that, far from exhibiting "zero-shot" generalization, multimodal models re
180.
by thesz 9mo ago
> But we shouldn't pretend that you will be able to do that professionally for much longer. Who are "we" and why do "we" "pretend"? > Will they ever carve furniture by hand for a business

More ›