code51

To counter that, imagine police bodycams with embedded AI touching up live-recording so that the police was always right, suspect always did something wrong. If you build the source device with this touch-up capability, if you are the certificate authority about these, if you are also defining how these recordings would be used in court, then a perfect fake is useful. (None of us see the actual sensor output from our mobile phones)

TL;DR In order for a fake to be useful, you have to fake AND be the only reputable source. If other recorders cannot get close enough, then your touched-up "real" video is the only material available.

Logically, shouldn't there also be a study taking a look at heart problems and blood pressure levels in people with congenital deafness?

We would then be able to separate the effect of noise vs. life stress. Surely, for any study about stress, socioeconomic status should also be a factor.

The idea that these little experiment wins bring user satisfaction and/or revenue increase is a pure assumption.

Employees get so aligned with these experiment systems that obvious faults in statistical approaches go unnoticed. At the end of the day it becomes just a way to put something on your promotion case.

In most cases, core product stays the same over the years without significant innovation.

Shouldn't there be an "invisible hand" at work to settle this problem automatically when Fed over-raises or under-raises? With the invisible hand and free market arguments, this should have been a non-problem at the first place. But... since the initial move was not natural (lots of cash injection), the natural final move has to be sudden and forceful. These analysis-paralysis rate hikes seem like lots of pawns to be lost before the final blow. It just opens a window of opportunity for ahead-of-the-curve retirees to save their wealth, not helping to avoid the final effect.

Yes, I did. I'm not denying the validity of the case here. My concern is related with the core risks of the procedure.

Timely colonoscopies save lives.

However, I have a hunch that frequent colonoscopies are maybe not a good idea.

Doubts:

1. We know more but we don't know all things creating colon cancer. Knowing the existence of cancers caused by viruses, disinfection procedures applied on the device could matter a lot.

2. Even if we disinfect properly, we know of a notorious little fellow: H.pylori, which can go down to the intestines. "Helicobacter pylori infection is associated with colon adenomatous polyps detected by high-resolution colonoscopy." https://bmcgastroenterol.biomedcentral.com/articles/10.1186/...

3. H. pylori (which serves just as an example, there can be other bad guys) is really hard to get rid of with disinfectants. It builds up resistance over time so it's doubtful that our disinfectants currently in use will stay as effective as well.

4. Frequently putting a foreign object to different people's colons (yes, I know it's disinfected) would increase the probability of transmitting such unknown risks.

I didn't talk about perforation risk. Just the possibly transmissible cancer risk.

So I remain doubtful that it's completely harmless It's necessary but every year? Doubtful without more data.

Finally, thanks for the people who invented and use it to save lives. I wanted to share my doubts on frequent colonoscopies

Developing anything around Custom Search API does not make sense because of the limits.

With GPT-J, search prompt works better if you don't try to get a JSON response. I don't know why but when you ask for technical formats/outputs, GPT-J heavily focuses on composing a technical document/code for you. It tends to start composing Javascript etc.

I'll come up with a pull request.

I set the base and engine in openai-api for goose.ai completions return fine, I'm seeing them fine in the log. However they are unusable to make Goopt work. The end of search prompts mock JSON format.

I get this \"content\": [{\"name\": \"Keto Diet\", \n \"typeClass\": {}, \n \"description\":[],\n \"contentImageUri\": null }]}

GPT-J doesn't seem to compose similar broken JSON so formatResults returns empty.

I made it work but GPT-J doesn't respond in the same manner to the template prompts for search. Goopt cannot use and display GPT output from there as a search result.

Apple Pencil might be the most fragile product Apple ever produced. It's like a ticking time bomb. Its battery is doomed to die any second.

I think that there can be a significant difference between every batch of coffee. Both the origin and conditions of the specific land where coffee grows matters in my opinion. People often imagine this as a factor that only changes flavor. I suspect the soil and the ecosystem (e.g. nearby plants, rivers) can trigger such negative effects even for an individual who's not sensitive to caffeine and consuming in moderation.

Looks like Getir has proven the business model locally but considering their latest moves in the UK, their main goal is capturing the European market. I have doubts about the whole instant-delivery space though. Profit margins are probably not large enough to make the business sustainable. The scenario looks like winner-takes-all or at least hold your breath until IPO.

Bluetooth headphones are a nightmare when your computer is under heavy load.

Hi there. I installed this and here's some feedback: I can't have any idea of the app because of my location. My experience can be summed up as: cold start. Why are you limiting the user with GPS location? Maybe I want to get an idea about a place I dream of traveling. Within: 15 km cannot satisfy this need. Plus if you suggest me to try "More: Within 15 km", maybe you can show a single-tap button to do it? I feel it's tiring to read the small text, understand it, don't know it (since I didn't go to preferences screen yet) and there update and come back. Bad UX.

Congrats for the launch.

Isn't there a "no free lunch" about remembering as well? Yes, memory training is probably a good idea but how do you pick what to store there in the first place? Forgetting things -properly- is an underappreciated skill in my opinion.

"What do we do with this information?"

Tell the lab to stop gain of function research.

Any kind of gain of function research lab needs to run under continuous international supervision from now on.

International Space Station (ISS) could serve as a good analogy here: have scientists and auditors from many countries in there so that we know what goes on day-to-day.

After that? Hold any country doing independent gain of function research responsible (from potentially building a bioweapon)

I'm fearing this will end up with a massive debt on their part.

What about D3K2 supplementation (adding in K2 supplementation) to counter these high levels of calcium?

begs for it: https://imgflip.com/i/3q1evo

NextDNS is great. However with these tools, problems appear when you specifically want to display your Firebase dashboard etc. directly going to a service from the "tracker" itself = analytics.google.com

Marking at domain level is too generic to wrestle with this problem.

Someone noted (vedtopkar) these are "HHV-6 and HHV-7, which cause roseola". Note that it's an incredibly common infection in babies.

There's also another study investigating association between HHV-6 and measles virus serology and brain antibody in autism:

https://www.ncbi.nlm.nih.gov/pubmed/9756729

Reading these studies makes me think an ineffective blood-brain barrier or a brain-specific defence might trigger common viruses to present a wider family of brain-related diseases.

We also know there is literature on encephalitis caused by both HHV-6 and measles. I'm speculating that even in a micro, benign form, this condition is enough the shake brain structure.

We should probably investigate a more complex but more generic model of brain diseases.

I think most people understand the benefits and greenness of nuclear compared to other forms of energy generation.

Yet it is also superficial to plainly say "nuclear is safe". That's obviously more complicated than that. Chernobyl, Fukushima (widely known, 2011), Marcoule (2011), Ibaraki (1999), WIPP (2014) etc.

These are not events like "Uber car killed a person but self-driving cars are the future" thing, these events are actual evidence that human error plays a huge part in dealing with nuclear systems.

Plus, when you think about the waste, it doesn't seem so green anymore. Maybe greener when you're throwing radioactive stuff into space.

What I'm saying is, it's not a simple issue that "average people don't understand" Actually we see that average people pick most pragmatic options when economically pressed.

Regulations are there for a reason. Personally, I don't want spacecraft throwing radioactive material onto our own atmosphere.

Imagine a higher culture formed in a super-earth, keeping peaceful with the earth and not leaving their home without a sufficiently eco-friendly technology being discovered.

They are used to their gravity.

It is hard to leave so they find value in sustainability earlier on.

They only leave with super-advanced form of green technologies.

If by "theory" you mean "darker skin in higher altitudes -> low vitamin D", I don't have anything against that.

If by "theory" you mean "low vitamin D -> autism", you're reading science wrong (by only using correlation when reasoning about it). You have to control for other factors in Somali population. That's how you do proper research. In its current form, as the researchers themselves point out, this says that a future study is needed.

Let me repeat: you might be entirely right but "this correlation supports the theory better" is not scientific evidence. It's just a hint to direct research.

As I said, if they find Swedish mothers with really low vitamin D and compare with Swedish mothers with high vitamin D, that would say a lot more.

I'm starting to think a study focusing on Swedish mothers with really low D (similar to Somali mothers) with Swedish mothers with high D would solve all the mystery here.

I wasn't trying to argue against. Just trying to understand better.

1. Yes, Somalis in Sweden have very low D. That is quite obvious [from Table 1] after winter.

2. Yes, Somali children born in Sweden have higher rate (according to this study: 17 in 2437 [0.7%] for Somali background, 484 in 250565 [0.19%] for non-Somali): http://onlinelibrary.wiley.com/doi/10.1111/j.1469-8749.2008....

However we should also consider the effect of other compounding factors about parents coming from Somali.

Here: https://link.springer.com/article/10.1007/s10803-017-3232-z

"Results lend support to previous studies that found higher rates of ASD intellectual disability in children of immigrants from low human resource index countries compared to other groups."

coming from Somali also means coming from a low human resource index country. all kinds of air/water pollutants and chemicals/pesticides might be at play -together with- vitamin D deficiency.

Vitamin D effect itself is hard to isolate from just this. Further studies need additional data about these other factors (air pollution, exposed chemicals etc)

Looks like difference of vitamin D levels were not significant in the 2010 study between mothers of Somali origin with or without a child with autism. http://onlinelibrary.wiley.com/doi/10.1111/j.1651-2227.2010....

Do you have another source for the statement in the first paragraph?

Negative aspects:

- Twitter auth requests the ability to send posts, an easy no

- signing up with email, you require an avatar picture. just use a placeholder

- can't test drive anonymously

If Facebook had quit providing any sort of value to the people, had quit providing incentive to store their valuable assets like photos, videos, opinions etc. on their system, then yes, I think if someone copied Facebook %100 and started operating it properly, they'd be a good social network.

Facebook is not just a code or just a network. ICQ had a network. MSN Messenger had a network. Heck Google actually has a network.

But without providing value, a network is not worth a thing.