2,130 karma · joined August 29, 2016
There are no Indigenous People with fancy believes systems taking care of Chernobyl Exclusion Zone, and the area is doing better than ever in terms of biodiversity despite nuclear disaster. The whole "magic" was to get civilization out of there.
https://en.wikipedia.org/wiki/Effects_of_the_Chernobyl_disas...
According to Wikipedia on Indigenous peoples:
"culturally distinct ethnic groups who are native to a place which has been colonised and settled by another ethnic group"
So, I guess, we would have been, if e.g. Germans had won the second world war.
The issue is that obtaining this understanding requires a lot of effort.
OpenAI invested in trying to understand how a simple CNN works.
You can read their results here: https://distill.pub/2020/circuits/
> In the original narrative of deep learning, each neuron builds progressively more abstract, meaningful features by composing features in the preceding layer. In recent years, there’s been some skepticism of this view, but what happens if you take it really seriously?
> InceptionV1 is a classic vision model with around 10,000 unique neurons — a large number, but still on a scale that a group effort could attack. What if you simply go through the model, neuron by neuron, trying to understand each one and the connections between them? The circuits collaboration aims to find out.
1.
> Do we have a retired BBC reporter that was based in an East European country during the Cold War? > (...) None of the above challenging semantic understanding functions can be ‘approximately’ or ‘probably’ correct – but absolutely correct.
There already exists systems that do text to sql translation. The question author uses as an example is actually quite easy, and not where modern text to sql systems tend to fail. The "absolutely correct" part is clearly false. Both humans and ML systems will make mistakes and have to deal with ambiguity. As a software developer working with nontechnical stakeholders, I'm 100% convinced that people routinely compose sentences they do not understand in any absolute sense.
2.
Good way to check if your proof is correct, is to check if it leads to absurd conclusions. His proof related to learnability (ML) and compressibility (COMP) would seem to indicate that either people can not learn or they do not understand language. Absurd.
3.
> In ML/Data-driven approaches there is no type hierarchy where we can make generalized statements about a ‘bag’, a ‘suitcase’, a ‘briefcase’ etc. where all are considered subtypes of the general type ‘container’. Thus, each one of the above, in a purely data-driven paradigm, are different and must be ‘seen’ separately in the data.
This is objectively false. Even simple context free embeddings like Word2Vec will capture some relationships between a ‘container’, a ‘bag’, a ‘suitcase’, a ‘briefcase’. Based on that alone the rest of that argument falls apart. But large language models go beyond context free embeddings.
4.
The last argument is relatively good one. It's hard to extract all attributes of human concepts from text alone. This is where large language models fail miserably. We need to provide models access to different modalities, but not only. The models can not be embedded in a static world of recordings in order to create rich understanding of things like agent or agency etc.
This is indeed missing, but at the same time this is being actively worked on.
So, that's it. Author makes 3 either false or absurd arguments. And one that is good, but wholly unconvincing as to impossibility of ML solving natural language understanding.
The same way as data scientist position has mostly nothing to do with publishing in peer reviewed venues with high impact factor.
Yes, of course! Is it the best way to spend your time having the goal in mind? Maybe not.
The issue with "learning math" is that there is more math being created than you are able to learn.
"a good computer scientist" is very broad.
Does it mean "a good 3D engines programmer" - then probably you want to know linear algebra, things like quaternions etc.
Does it mean "a good browser front end developer" - then probably you want to invest your time somewhere else, maybe web design, learning how to sketch.
https://ec.europa.eu/clima/policies/ets_en
https://en.wikipedia.org/wiki/European_Union_Emissions_Tradi...
Sci Hub seems to not fetch it.
> Tesla (...) are doing much worse than those ordinary drivers.
That was response to:
> Those successfully tested ordinary drivers also kill many people every year.
The not-apples-to-apples comparison indicate that Tesla are twice less likely to cause accident/crash when on Autopilot, and even less when not on Autopilot.
I find it also highly likely that Autopilot slightly increases the chances of accident right now.
SpaceX is still cheaper, has some 10 times more capability of cargo delivery to the Moon and working hardware.
NASA will never agree to what Blue Origin is proposing, unless congress changes the law.
> By comparison, NHTSA’s most recent data shows that in the United States there is an automobile crash every 479,000 miles.
So, Tesla numbers are better even in the article you link, but still not apples-to-apples comparison. Accident vs. crash.
And Tesla is adding pro forma camera driver monitoring.
My brother bought me iPod as a present, it was cool device, but I failed to upload music to it.
On normal MP3 players I just moved files onto the player and it just worked.
Apple on the other hand... I never figured out how to do it with iTunes, I still don't know.
Some years later after installing Ubuntu I had an idea of using Rhythmbox to try to put music on that failed iPod and that worked like a charm.
Ubuntu instillation was easier for me than iTunes.
Well, I never even thought about buying Apple device ever since.
NHTSA is investigating Tesla, some 20-30 accidents. If they find that Tesla is doing something seriously dangerous, I'm sure they will force Tesla to take corrective actions.
The same applies to European regulators.
> Evolvability is defined as the capacity of a system for adaptive evolution. Evolvability is the ability of a population of organisms to not merely generate genetic diversity, but to generate adaptive genetic diversity, and thereby evolve through natural selection.
In changing environment organisms have to be able to quickly adapt.
You can compare organisms like Greenland Shark with lifespans of up to 400 years with a common mouse. Mice will go trough some 100-1000 generations in lifespan of a Greenland Shark. Which one is more likely to adapt to changes in the environment induced by humans and which will go extinct?
Anthropocene extinction event will eliminate a lot of life forms with long lifespans, many are already gone, and previous extinction events probably did it too.
BTW - This is a falsifiable hypothesis in a fairly simple simulation.
The default for humanity is ever-present death due to diseases, hunger and thirst.
In 1820 6 out of 10 kids would die before they would get to be 5 years old.
Imagine having 6 kids just for 4 of them to be dead before they got to be 5 years old.
Today poverty is the lowest it has ever been, basic education and literacy is the highest it has ever been [1].
If this is how the collapse will look like then I'm all for it.
I also recommend "Don't Panic" by Hans Rosling [2].
[0] https://en.wikipedia.org/wiki/Socialism_with_Chinese_charact...
[1] https://ourworldindata.org/uploads/2017/01/Two-centuries-Wor...
Radar throws a lot of false readings so there are many situations where it is ignored.
For examples, readings of stationary objects are mostly false.
My guess is that the speed difference was so big that also in this case Tesla decided that the radar reading was stationary and false.
Here is finding from one study that was peer-reviewed and got published in a journal "Vaccines" (impact factor: 4.4):
https://www.mdpi.com/2076-393X/9/7/693/htm
> For three deaths prevented by vaccination we have to accept two inflicted by vaccination. Conclusions: This lack of clear benefit should cause governments to rethink their vaccination policy.
The study is now retracted, but for one such outrageously wrong study there are probably 10 more that do not receive so much scrutiny.
There is nothing that can be done if you do not trust authorities, besides spending time on getting into the weeds.
But if you trust authorities, then according to American CDC:
> CDC recommends everyone 12 years and older should get a COVID-19 vaccination to help protect against COVID-19.
https://www.cdc.gov/coronavirus/2019-ncov/vaccines/recommend...
And according to WHO:
> Children and adolescents tend to have milder disease compared to adults, so unless they are part of a group at higher risk of severe COVID-19, it is less urgent to vaccinate them than older people, those with chronic health conditions and health workers.
> More evidence is needed on the use of the different COVID-19 vaccines in children to be able to make general recommendations on vaccinating children against COVID-19.
> WHO's Strategic Advisory Group of Experts (SAGE) has concluded that the Pfizer/BionTech vaccine is suitable for use by people aged 12 years and above. Children aged between 12 and 15 who are at high risk may be offered this vaccine alongside other priority groups for vaccination. Vaccine trials for children are ongoing and WHO will update its recommendations when the evidence or epidemiological situation warrants a change in policy.
> It's important for children to continue to have the recommended childhood vaccines.
https://www.who.int/emergencies/diseases/novel-coronavirus-2...
By analogy, you are basically using the word African to mean "a person in possession of someone else".
https://www.etymonline.com/word/slave
@edit The fact that people down vote this highlights that the whole issue is just virtue signaling.
If you start a business with Amazon it's probably a good idea to google other companies experiences.
But I agree that everything counts, including political aspect of individual actions.
We just can not feel good about ourselves because we don't eat meat, fly airplanes and avoid buying plastic. It's not enough.
It is beauty of the free market and capitalism that as the price of CO2 will increase everyone will start looking for ways to reduce emissions without even thinking about emissions. Including innovators and investors.
There is also nice political side of Emissions Trading Schemes. It had no effect when it was introduced and it's impact increases little by little each year. Not everyone who would be opposed realized effects of that system when it was introduced. They are slowly gaining awareness, but the positions have reversed. Now, they have to fight against the status quo.
I see that quite clearly in Poland. Polish coal miners have a lot of political swing, but they did not realize the consequences of Emissions Trading Scheme. Now as the effects are increasing Polish government is not capable of changing it. But they have money from selling emission rights so they can use that money to appease coal miners and other politically important groups.
> But why would systemic solutions get suggested or implemented if individuals didn't show they wanted a solution? Where would these initiatives come from if they're not influenced by the voices, voting power and buying power of individuals?
It's easier to convince people that climate change needs to be addressed than to educate them on how --all-- industries work. E.g. how much do you know about concrete curing, steel and related emissions? Would you be able to make informed decisions with regard to building materials and CO2 emissions? Do you even know if viable alternatives exist? Do you expect companies to spend more to emit less?
There are systemic solutions:
European Union Emissions Trading System is quite good system. It sets a limit on emissions and auctions rights to emissions that can be traded later. The limit will go down according to the IPCC recommendations. The current price is ~50 euro per metric tone of CO2 emissions.
https://en.wikipedia.org/wiki/European_Union_Emissions_Tradi...
Overall, charging really will not be an issue. Most people will charge at home or work.
And in 2025-2030 there will be plenty of chargers that will push 200-300 km in 15 min on longer trips.
1. Cheap
2. Takes very little space
3. Good for health
4. In a dense city almost as fast as a car; sometimes faster
For people who doubt biking is possible. In Finland there is a city near arctic circle called Oulu with population of around 100 000 people. They have 2h of day light in winter. Temperatures often go to -10 degrees centigrade. And 22% of all trips are done by bikes. Including children going to primary schools.