ke88y

19 ms·

181.
by ke88y 3y ago
There's a lot of wisdom here. Stupid but hard goals are still hard goals. Necessity -- not taxpayer largess -- is the mother of all invention. The corresponding risk is in the resulting cargo cult around arbitrary constraints whose rea
182.
by ke88y 3y ago
The US military as a guaranteed purchaser, maybe. The US military as an organization, definitely not.
183.
by ke88y 3y ago
> they'll probably be able to find a bank willing to charge them < $0.25 I guess that includes some very small values, but the upper bound seems ridiculously high.
184.
by ke88y 3y ago
LOL what? Keeping private keys private is not "security through obscurity". Or if it is then basically all security is security through obscurity. No one is posting their private keys on github, and when they do their crypto goes
185.
by ke88y 3y ago
FedNow doesn't allow any more or less tracking than what's already possible through the Fed's visibility into ACH. So for the purposes of FedNow the entire conversation is off-topic. Anyways, you missed the point. The anti-ne
186.
by ke88y 3y ago
At the very least, one would hope that credentials and perhaps also certain design documents such as threat models aren't public. There may also be implementation details or code which are subject to NDA, either from the Fed itself or
187.
by ke88y 3y ago
> I can buy an icecream anywhere in the world without anybody tracking me. No you can't. Companies like OpenEye and Deep Sentinel offer facial recognition solutions that are targeted at loss-prevention, and these systems are commo
188.
by ke88y 3y ago
The ISP duopoly/monopoly is worse from a consumer perspective and functionally equivalent from a civil rights perspective. To think otherwise is naive. In fact, the private sector solution is perhaps even worse from a civil rights pers
189.
by ke88y 3y ago
I view this type of thinking as the conspiratorially minded's version of "trying to solve social problems with technical solutions". If there were political motive and will to implement such a system, it could be done without
190.
by ke88y 3y ago
The end users of FedNow are financial institutions, so not directly.
191.
by ke88y 3y ago
Boston's income-restricted housing program is available up to 150% AMI, and 100% AMI is a pretty common cutoff. A 6 person household with $150K is 90% AMI. In fact, even $200K at 120% AMI still qualifies for some units. There's a
192.
by ke88y 3y ago
The worst part is there's no particular reason for this -- infusing proper software development best practices into existing EE/ME coursework isn't that hard. It's an instance of the larger pattern in which technical deg
193.
by ke88y 3y ago
I don't follow. Did you accidentally reply to the wrong comment?
194.
by ke88y 3y ago
Kids are expensive and housing is intractable where the mid-level jobs are. In some housing markets, a family with 2 boys and 2 girls earning $150K/yr will qualify for government subsidized housing. It's not 1995 anymore and not e
195.
by ke88y 3y ago
> I do a lot of support work for Control Systems. It isn't unheard to find a chunk of PLC code that treats some sort of physical equipment in a unique way that unintentionally creates problems. I like to parrot a line I heard elsew
196.
by ke88y 3y ago
As someone who worked on publicly funded grants, I find this attitude repulsive. Everyone who receives money from public coffers owes that public a service mindset. The arrogant self-importance in academic science is inexcusable and reason
197.
by ke88y 3y ago
> The showing up at 10Am and leaving at 3 with a long lunch and working 20 hour work weeks. Mostly this. The conferences honestly aren't that much of a perk, relative to the pay differential, at least in STEM fields. The "hot
198.
by ke88y 3y ago
This is why my teams always provide explicit opportunities and spaces for professional development. People should have the opportunity to stretch and grow; if you don't provide those opportunities explicitly, then your most motivated e
199.
by ke88y 3y ago
> What would you do besides the big fat red disclaimer we have in the project to address your concerns? I probably wouldn't use an LLM for this problem domain. Or, if I did use an LLM, it'd be as a way to map the user's
200.
by ke88y 3y ago
> Doesn't it seem like this kind of fraud should result in prison sentences? It depends on the nature of the fraud. I'm no fan of academia (see my post history), but this has to be close to the bottom of my list of priorities
201.
by ke88y 3y ago
Slight nit: in the case of MIT and other "IT"s the story is slightly more complicated. Shockingly few students who graduate with Bachelors degrees in the USA make it through even a watered down Calculus I, let alone MIT's mul
202.
by ke88y 3y ago
That's a nice narrative. How is any of it relevant to a system that uses LLMs trained on user-written reviews of health supplements to recommend products based on end user's descriptions of health conditions?
203.
by ke88y 3y ago
No, I get his point. I think the transmission of knowledge and insight and clarity works very differently for executable artifacts than for artifacts written in natural language. > This is not affected by having a better way to express
204.
by ke88y 3y ago
I generally agree. Until the SD and Copilot cases are settled, there's a huge risk that training is Fair Use.
205.
by ke88y 3y ago
> I don't think I need to be a SCOTUS judge or powerful Senator for those opinions. The world isn't fair. Unless you think you can imminently effect political or judicial changes, you need to ACT in the world according to ho
206.
by ke88y 3y ago
Your first two paragraphs are your opinion. Unless you're a SCOTUS judge or powerful Senator, I'm not sure that your opinion matters much. And I do agree with your opinion, btw, so you're in good company, but I think the cour
207.
by ke88y 3y ago
If you have a working LP solution without any glaring compromises in the problem formulation, then I'm not sure why one would want to throw out a perfectly good working solution... algorithms are the means, not the ends :)
208.
by ke88y 3y ago
PhD student and faculty (quite briefly). Serving on grant panels also informed this view, but looking back, I think I was already getting disgusted by the culture of academia around my second year of grad school. My immediate family spent&#
209.
by ke88y 3y ago
> how easy is it to know that a reinforcement learning solution is actually a global optima and not just local? If the problem is too complex for LP then you're probably not going to get a global optimal from RL either.
210.
by ke88y 3y ago
The cons are that RL sometimes requires more babysitting, more non-subject-matter (ie, optimization-method-specific) expertise, and can be a bit less... predictable... at inference time. Also, LP often offers perfectly acceptable optimizati

More ›