jmward01

53 ms·

961.
by jmward01 2y ago
I think a big missing point is Junior people force an org to write better, more obvious, code and more durable systems in general. "If you can't explain it to a 5 year old..." applies here. The other major missing point is th
962.
by jmward01 2y ago
I had a discussion with someone recently about how tech is enabling enforcement of laws/policies/etc at a scale never before possible and how that is a bad thing. The problem is that people can't conform perfectly and the tig
963.
by jmward01 2y ago
Soon mice will live forever, with 100% cognitive function, no hair loss and free of cancer. Now if only we could do the same for people.
964.
by jmward01 2y ago
I only use mine as a timer/clock and even that is becoming more rare. Every other feature is so add encumbered as to be useless. Echo is dying, quickly.
965.
by jmward01 2y ago
In his TED talk Weller points out that latitude correlates with heart disease well. This study gets to clear out a lot of variables by sticking to the UK though so it is really valuable in testing the potential of UV without having to deal
966.
by jmward01 2y ago
In his TED talk, Weller gives a potential alternate benefit source in nitric oxide production which leads to potential cardiovascular benefits. So, the short answer is it may not be (just) vitamin-d. The bigger point though is that there ar
967.
by jmward01 2y ago
He has a good TED talk on this from over a decade ago. https://www.ted.com/talks/richard_weller_could_the_sun_be_go...
968.
by jmward01 2y ago
Bias and assumption are needed features of non-deterministic thought. We live in an infinite universe so thinking and exploring until you have truly 'got it right' would lead to death. I have built non-deterministic NNs (which are
969.
by jmward01 2y ago
True. It would be interesting to run these same tests on the 7B model to see if trend information changes or not. 7B had a march cutoff so if the aug-dec dip migrated to oct-march (or just disappeared) it would be strong evidence for traini
970.
by jmward01 2y ago
I wonder if the dip is more about LLama3 70b training and data than a change in sentiment. The data cut-off was Dec 2023 for 70b. That looks to coincide with the reversal of the dip.
971.
by jmward01 2y ago
It sounds like an independent third party needs to review what happened.
972.
by jmward01 2y ago
I want all of this and it really shouldn't be hard. In fact, many have tried at least parts of this but failed because of the regulatory hurdles. Taking 10 years to bring a product to market while spending a huge amount of $$$ testing
973.
by jmward01 2y ago
Aviation discussion starts at the 9:00 mark. Lead-up discussion lays the battery roadmap that has gotten to this possibility.
974.
Battery powered flights from Washington DC to LA. No longer a pipe dream?[video](youtube.com)
2 points by jmward01 2y ago|1 comments
975.
by jmward01 2y ago
This is a bigger deal than the 'degraded' implies. SQS has basically ground to a halt for reads which is leading to massive slowdowns where I am at and the logging issues are causing task timeouts.
976.
by jmward01 2y ago
Or maybe it is just over-reporting. Aviation is safe according to the data, but not according to the amount of reporting given to every minor problem. As this story pointed out one of the many backups, ATC, called out that they were low bef
977.
by jmward01 2y ago
I never thought I would say this but thanks Meta. *I reserve the right to remove this praise if they abuse this open source model position in the future.
978.
by jmward01 2y ago
Nature has one solution to creating durable complex systems. Create an ecosystem with a lot of diversity in it. Mono-anything is bad because of all the obvious reasons that we have just seen.
979.
by jmward01 2y ago
I cited several examples of value and I never said earthquake proof, I said preparation. Improve roads and critical services. If you don't see value there then I think we are done with the discussion. Have a nice day.
980.
by jmward01 2y ago
Companies are evil when they lobby to change laws in their favor, not when they take advantage, to the maximum extent possible, of the law. Just taking advantage of the law is rational, it is changing it that makes them evil. In this case t
981.
by jmward01 2y ago
Sounds like you have just as strong a misconception of people as the people you are trying to call out. The thing about big cities is they draw a lot of people from everywhere. The diversity of experience is high so you may be surprised abo
982.
by jmward01 2y ago
I'll grant I could have said 'In the long run requiring better infrastructure probably won't hurt the economy and likely will actually help it.' instead of giving the focus word 'buildings'. Past that, I think
983.
by jmward01 2y ago
I think many of the discussion threads here can be summed up as 'you can't tell me what to do' but they don't acknowledge that when people live together everything done, or not done, impacts them all. Appropriate infrast
984.
by jmward01 2y ago
There will always be mistakes and corruption but it shouldn't stop us from doing things that in general pay big dividends. In the long run requiring stronger buildings probably won't hurt the economy and likely will actually help
985.
by jmward01 2y ago
Probabilities are very hard to think about so I personally relate back to driving. We do it every day but admit it is dangerous and take precautions like requiring testing and licensing, regulate the manufacture of vehicles and inspections
986.
by jmward01 2y ago
The economy isn't a zero sum game and infrastructure projects generally pay for themselves in the long run many times over. I think we are too interested in today's budget to pay attention to historically valuable things.
987.
by jmward01 2y ago
What does it take to convince people to take risks like this seriously? The challenge is that hypothetical risks 'sometime in the next 50+ years' don't convince people to do things, at least in the US. The pacific northwest i
988.
by jmward01 2y ago
yep. I wrote without thinking. c^n would be better right.
989.
by jmward01 2y ago
The fundamental issue isn't addressed but it sure is hinted at. Scheduling is an NP problem. As queue size, or backlog, or (insert thing here that tracks work to be done) grows it takes n^p calculations to schedule it optimally. This i
990.
by jmward01 2y ago
> and self-serving analogies to human learning should be rejected. Not sure how this is self-serving or a reference to human learning. This is a pretty emotional reply and it makes it hard to see the points you are trying to make. I thin

More ›