I spent 1 month around Kyoto's UNESCO sites 10 years ago while working remotely for a company in the US and am not sure I'd want to spend whole 6 months in Japan when there are so many interesting places all around the world. One needed to know basic Japanese as very few people spoke English, finding an accommodation was a chore as AirBnB was non-existing and hotels with unlimited fast WiFi scarce and one had to book an apartment in some of their temporary housings months in advance. Overall a nice and unique country though people looked overworked (young people complaining about working 996 lifestyles) and isolated from foreigners, similar to Germans (except their children that always wanted to practice Engrish with anyone not looking Japanese).
The author might be right though what I've noticed with DL models is that the theory is often leading to underwhelming results after training and "bugs" in models sometimes lead to much better real-world performance, pointing out some disconnect between theory and what gradient-based optimization can achieve. One could see it also in the deep reinforcement learning where in theory the model should converge due to being Markovian via Banach fixed point but in practice the monstrous neural networks that estimate rewards can override this and change the character of the convergence.
Concept of time and eternity - we are in time that flows in one direction, everything has a beginning and end. A being in eternity outside time views all moments of time all at once (like all individual pictures of a movie), hence no origin is necessary/defined in eternity.
iPhone used capacitive touchscreen whereas the competitors used resistive touchscreens in their smartphones, which instantly added more usability to iPhone compared to Nokia N95/97 that was already a fully featured pocket computer in mobile case and likely much more powerful than the original iPhone. Apple did the dirty logistics trick on other smartphone manufacturers by buying all production of capacitive touchscreen factories and similar key components 3 years ahead, leaving other phone manufacturers unable to respond.
I personally think MSFS 2020 and Automobilista 2 are killer VR apps. All (wannabe or pro) pilots could learn how to operate any given aircraft in MSFS and relive past glorious racing in A2 on current or historic tracks/cars. Senua and Alyx showed what is possible in gaming as well and why it's so much better than 2D. Elder Scrolls looks great in VR just the controllers make it a joke when fighting (too easy and weird). I still think 4k is too low and 8k will be needed to feel like a 1080p phone.
Finding anything valuable in vacuum robot video feed might be similar to trying to find a needle in a haystack - who has the capacity to go through billions of hours of videos and analyzing if there is anything worth attention? Both companies and hackers would go bankrupt pretty quickly.
We are talking about a vacuum cleaner that is 99% of the time parked in some closet. What kind of valuable information are they going to get? Now compare it with Amazon Echo, your TV with a webcam/microphone, your smartphone - those are far higher risks.
How is a company 10,000 miles away from you going to hurt you compared to a local one that has access to all government services (and vice versa)? I'd rather prefer Chinese or Russian companies to spy on me because what are they going to do with it that would matter? Will I get more spam in Mandarin/Cyrillic I can't decipher anyway?
When I was suffering from this I tried to hack it and what I noticed was that when I got more blood to the brain (e.g. by being upside down), it improved (for a minute or two). So that led to the brain hypoperfusion theory and then I decided to ramp up the red blood cell count to give my brain the ability to do more with less, by taking liposomal iron, lactoferrin and liposomal sublingual B12. This led to a massive boost that I could work normally for 8 hours (but had to take rest afterwards).
It's one datapoint. Not sure why I should by default distrust somebody whose life was ruined by it and is essentially involved in a human-style reinforcement learning crowdsourcing trying to figure out how to recover by sharing what worked and what didn't.
More like with neuroinflammation and brain hypoperfusion? Both present and measured in long covid sufferers. M1 polarized microglia, damaged endothelium, dysfunctional mitochondria...
There was some recent study where most patients who received monoclonal antibodies regressed rapidly so I am not sure that's a good pathway. The hypothesis is that the covid antibodies are quite nasty so the body creates anti-antibodies that in turn resemble the original virus wrt certain effects and one ends up in a cycle of doom that first needs to be broken before one starts getting better.
I did 35 days on electrolytes/vitamins-only. The first two days were annoying, then it was a smooth sailing for a week (did two bike climbs at the end of the week) and then gradual decrease of physical energy/abilities while mental energy stayed constant. At around three weeks some feelings of euphoria arrived, then a week later the opposites. I was thinking whether to continue or not, but I reached my target weight (losing 1lb/day) and my weight loss slowed to less than 0.5lb/day so I called it quits. Refeeding was kicked off with a high-dose thiamine and some glucose to avoid refeeding syndrome, then I spent a week on bone broth, slowly adding more stuff in.
A few months later covid arrived, I got it and am not sure if the fast helped surviving it or actually worsened it as while I had almost asymptomatic run, in a few months it morphed into long covid. I think I depleted my B vitamins during the long fast and kicked off mitochondrial dysfunction as a consequence of covid.
"We receive a lot of applications" can be also a marketing speech and it could also mean they are flooded by spam requests from all over the world they can't filter out.
Let's just make qdrant to pinecone/weaviate/redis/etc. data exporters, that would make the company super happy! Free labor benefiting their competitors.
Cognitive Search already contains hybrid search (vector + BM25 + custom ML reranking) and they use chunks of 2048 tokens with a custom tokenizer. So it should be now better than most vector DBs. One could probably make something better by using some version of SPLADE instead of BM25 but their secret sauce lies in their custom ML model for reranking that gives them the largest search performance boost.
Why is AGPL bad? It would prevent Amazon from taking it from the founders and making money off it without giving anything back, like they did to dozens other products.
It's likely due to being close to Vienna that would take an indirect hit in that case. Also, it seems to be covered in the follow up fallout so it would be severely damaged anyway.
WW3 or some other high-intensity conflict could be potentially launched by highly coordinated deepfakes. Imagine creating some videos offending an aggressive religion with many combatants or generating event videos that didn't happen and managing narrative as the uproar progresses etc.
I know headhunters that specialize in identifying highly functioning sociopaths for leadership roles. But experiencing total dehumanization of others first-hand and subsequent rise of that person is something one never forgets.
My very first boss that I believe is an undiagnosed psychopath is now director level over there. That told me everything I needed to know about any possible direction of the company. I no longer use anything made by them and actively steer everyone I know away to alternatives.