nyrikki

41 ms·

541.
by nyrikki 2y ago
Enterprise Architecture...which most people still don't understand what it is even with the full term :) Typically it is thought of as a Clinger-Cohen Act compliance task and/or policy enforcement...I am in the enabling team auton
542.
by nyrikki 2y ago
Pi is transcendental but computable. For an introduction to real real analysis is an old book typically called the 'Baby Rudin' Principles of Mathematical Analysis by Walter Rudin. It is tiny and expensive, as most college math te
543.
by nyrikki 2y ago
If the goal was to create an artificial neural network that better approximated the biological human brain, yes the perceptron model is insufficient. If your goal is to produce a useful model on real hardware and it works...no Remember the
544.
by nyrikki 2y ago
Easiest way I can figure out how to explain my claim. Consider when you use 'partial connectivity', E.G. convolution or pooling layers for local feature extraction on say MNIST. While useful, those partial connection layers are ex
545.
by nyrikki 2y ago
This whole situation is been a real mess for years. 1) Quiet quitting is primarily the result of disengagement. It is typically an organization failure and not an IC problem. 2) More than a fifth of the workforce was laid off early in the
546.
by nyrikki 2y ago
Yes sorry.
547.
by nyrikki 2y ago
Unfortunately dendritic compartmentalization, spike timing etc are still not present. All efforts at models of SNNs that I know of have hit problems like riddled basins so far, that is what to look for to move past the limits of perceptron
548.
by nyrikki 2y ago
You need to move past decimal expansion, there are sufficient methods to prove one sided limits on the reals in finite time.
549.
by nyrikki 2y ago
A Turing machine as an infinite long tape, and even if it takes 10 times the time to the heat death of the universe x<y will halt. Finite time is not practical time. Given two Turing machines A and B approximating numbers a and b, where
550.
by nyrikki 2y ago
A computable number is real number that can be computed to by a finite, terminating algorithm. x<y meets that definition because you can compare digits until one doesn't match. The same can be done for not equal, testing until there
551.
by nyrikki 2y ago
Equality of real numbers is undecidable. Specifically recursively undecidable, which is the same as uncomputable. The Real numbers are topologically weird when you dig into it. Random example. https://mathworld.wolfram.com/
552.
by nyrikki 2y ago
The platen and paper feed stops line for line. the drum or similarly the chains used on the hammer printers were in constant motion. I am just old enough that I had to repair both types at the beginning of my career. Although typically reb
553.
by nyrikki 2y ago
Halt is a decision problem, TFNP is a combinatorial problem constructed of decision problems. The HALT decider is a total Turing machine that represents a total function. I am struggling to explain this in a rigorous fashion, and there are
554.
by nyrikki 2y ago
> In startups, effective team structures are usually functionally organized. Startups work with functional structures because they are small enough to not have communication or political impediments. The structure simply doesn't hav
555.
by nyrikki 2y ago
As you are talking about formal proofs, and not the scientific counterexamples modern programming uses: Proving a function is total in the general case is a NP total search problem. IIRC this is equivalent to NP with a co-NP oracle, or the
556.
by nyrikki 2y ago
While imperative, and not 'pure' even C was created to set upper bound of the number of iterations of every loop being known before entering, thus PR Dennis Ritchie's research at MIT was focused on what he called loop program
557.
by nyrikki 2y ago
This result could happen without any intentional conduct, so ironically you may have made a similar error as the original researchers. While there are very real issues about reproducibility and motivation, rarely do studies actually claim w
558.
by nyrikki 2y ago
Fleets don't go to jiffy lube, when a car isn't being rented (e.g. at the shop) it is costing the company money, and Tesla's price drops cause issues, as covered in Hertz's earnings call. https://ir.hertz.com&
559.
by nyrikki 2y ago
The challenge of repairs especially from gigacasting is almost certainly a problem at that scale, even without Tesla's slow parts chain. After finding damage in a car I was going over for a friend, I would avoid them personally even if
560.
by nyrikki 2y ago
This video (I know sorry) will help out with dimensionality in many computing problems, which isn't the concept here, or what many people think of infinite or high dimensionality. It is worth your time IMHO. https://youtu.be
561.
by nyrikki 2y ago
While not explicitly about your chosen tool chain, monetizing user data is absolutely going to happen in the LLM space. While there may be some holdouts, Wall Street is already questioning the payback and as some big players are over promis
562.
by nyrikki 2y ago
The shear number of apps that provided no value to the users, but were merely a data grab has resulted in a serious impediment in my circles. I think the industry is going to need to find a new source of income or focus on customer value be
563.
by nyrikki 2y ago
Besides a dog whistle for some and a fear lever for others? > Hatred is the most accessible and comprehensive of all the unifying agents. Mass movements can rise and spread without belief in a god, but never without a belief in a devil.
564.
by nyrikki 2y ago
It is part of an effort that will have impact. Remember that the US federal government is the largest IT spender in the world. This is part of their efforts to move agencies internally I using a carrot first. Now that the GAO officially re
565.
by nyrikki 2y ago
For moral reasons, historically I never wrote POCs or threatened disclosure. For companies like Microsoft, which a CSRB audit showed that their security culture 'inadequate', the risk of disclosure with a POC is about the only too
566.
by nyrikki 2y ago
Cheese and Chalk. It is very different to generate synthetic datasets to assist in targeted training , vs ingesting LLM output from web scraping.
567.
by nyrikki 2y ago
It is more about vendor and technology mitigation vs selection. Companies fail, projects degrade, like lask week quality can go down. The better question is why would my organization couple their success to you wagon, do you provide them wi
568.
by nyrikki 2y ago
Physical security is probably a bad example, when people need physical security they end up calling installers. ADT got where they are by making their installers lives easier, and make it easier to find those installers. I am not the market
569.
by nyrikki 2y ago
The Mars Climate Orbiter, which had a problem with units from a vendor, but ultimately crashed because the people who noticed the problem were ignored.
570.
by nyrikki 2y ago
Prediction: The culture issues that are almost certainly the ultimate cause will not be addressed, and only tighter controls will be considered. This despite major government failures like the Challenger accident, Vietnam, or the Mars lande

More ›