Slashdot Mirror


User: Visarga

Visarga's activity in the archive.

Stories
0
Comments
246
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 246

  1. Re:only 2 ports and no power jack? on Apple Rumored To Remove Old-School USB Ports On Next MacBook Pro (vice.com) · · Score: 1

    Deep learning and scientific calculation are also dependent on nVidia GPUs, and I seldom hear anyone suggesting Apple for that use case. Maybe a MacBook for SSH.

  2. Re:Because Windows Sucks on OMGUbuntu: 'Why Use Linux?' Answered in 3 Short Words (omgubuntu.co.uk) · · Score: 2

    > The Linux ecosystem is growing, and getting more an more useable as a professional platform

    The Linux ecosystem (or more exactly, open source) is a black hole that eats voraciously and has become so big that Windows is just including it as a subsystem. They can't block or ignore it any more. The more software is contributed in open source, the more powerful attraction it has.

  3. Because of community on OMGUbuntu: 'Why Use Linux?' Answered in 3 Short Words (omgubuntu.co.uk) · · Score: 1

    You should have stopped at "because of community". Where is the ethos, there is the best place to develop. You get to benefit from the cumulative creativity of many of smart people.

  4. What these "symptom checker apps" on Doctors Perform Better Than Internet Or App-Based Symptoms Checkers, Says Study (sciencedaily.com) · · Score: 1

    I didn't catch the name of the apps they evaluated. In machine learning, the difference between cutting edge and older algorithms could be very large. If they compared AI to humans, they should at least compare to the best medical diagnosis AI.

  5. Re:Smartphone? on Samsung Ships Flameproof Boxes For Note 7 Returns (arstechnica.com) · · Score: 2

    Beep Blup Bop! I am a bot. I will send a message to the user "Anonymous Coward" when somebody invents a smartphone.

  6. Re:Uninterested in a personal relationship on Google Hires Joke Writers From Pixar and The Onion To Make Assistant More Personable (cnet.com) · · Score: 1

    I want it to understand context more than the past 2 lines of conversation, I want it to have a common sense (try "What is heavier, a dog or an elephant?" and see if it knows), I want it to do simple reasoning, to be extensible with APIs and use scripts, and yes, I want it to make pointless conversation and not default to web search every time conversation falls outside its knowledge graph. Maybe it could run a web search and extract the relevant information from it into a text reply. Assistants shouldn't need to default to web search.

  7. Making a more personable assistant would analogous to eye candy in web design. A certain amount of it is useful.

  8. Re:The Biggest Joke of All on Google Hires Joke Writers From Pixar and The Onion To Make Assistant More Personable (cnet.com) · · Score: 1

    > The idea that anyone would trust Google with a microphone always listening in their living room itself feels like an article from The Onion. The idea that anyone would entrust their web search logs to be collected by one company ... seems more dangerous.

  9. Re:Yeah, that's how I want to spend my free time on YouTube Is Looking for Volunteers To Improve Its Site (fortune.com) · · Score: 1

    YouTube is a treasure trove of music, courses, hobby videos and news. It's not dying, and if it did, it would be a tragedy.

  10. So companies can't sell gun designs on With 3D Printer Gun Files, National Security Interest Trumps Free Speech, Court Rules (arstechnica.com) · · Score: 1

    If a terrorist wants to print a gun, is he going to mind that he is breaking the law? This law doesn't protect us from terrorism. Once the capability to print guns is there, no amount of legislating is going to prevent anyone from getting some gun designs form the black market, or from the internet, or even design their own. Pandora's box has been already opened.

  11. Re:Not a nice way to die on How Cities Are Using Dry Ice To Kill Rats (usatoday.com) · · Score: 1

    > We better hope we don't encounter aliens that feel the same way about us and see us as pests on "their" new planet.

    We'll probably not encounter aliens soon, but the probability of developing AGI is much higher. I hope it sees us as cute pets, not rats.

  12. So, the solution is to burden users with choices? on Stanford Engineers Propose A Technology To Break The Net Neutrality Deadlock (phys.org) · · Score: 1

    So, the solution is to burden users with choices? How is an uneducated person or a child going to shape her internet traffic? They have to learn that the internet is not just a big truck first.

  13. Re: Horribly bad and confusing summary on Google's DeepMind Develops New Speech Synthesis AI Algorithm Called WaveNet (qz.com) · · Score: 1

    But this time the source article is really nice. DeepMind's blog is quite good.

  14. Re:I'm unimpressed on Google's DeepMind Develops New Speech Synthesis AI Algorithm Called WaveNet (qz.com) · · Score: 1

    Nah, I still prefer Alex from Mac OS. The IBM voice is smooth but unnatural in intonation, even in the example where they marked intonation on the text. I really loved the DeepMind samples, but they come at 1 second of speech generated in 90 minutes of computation, so, no chance of having that voice on my laptop.

  15. They pulled a Microsoft.

  16. Re:No benefit other than losing the cord on Apple Removed Headphone Jack From New iPhones Because It Owns Largest Bluetooth Headphone Company (theverge.com) · · Score: 1

    The new apple wireless buds don't over any option for people who prefer full headphones covering their ears.

  17. You can fumble with the adaptor if you are out of juice in your wireless headphones, of just buy two sets of wireless headphones and solve the problem. /s

  18. > Basically it's a cash grab.

    They only stand on one leg: bluetooth is awful in practical usage. The time and difficulty of connecting and disconnecting it makes it almost not worth the effort. If this wireless standard takes off, then there will be the justification for Apple's self serving move.

  19. Some people say "just connect it at the end of your headphones cable and forget about it". But what if you want to use your headphones with your laptop later? Then you have to take off the dongle, store it somewhere safe, and later find it for when you're going back to your phone. So you have to mind another object - where is it? did I left it on the cable or put it somewhere else? Being such a small object makes it easier to lose. Being a white thing, it looks bad at the end of a black headphone cable, especially if you're proud of your headphones.

  20. But there is a solution, and Apple agrees with it wholeheartedly: buy two sets of bluetooth buds. They charge in their carrying box, so when one set discharges, just swap it with the other set. Also, it's going to help when you start losing pieces of it, you're still going to have one set.

  21. The user should assume to pay a sum of money, say, 20$ per month, in total. The actual tracking should be implemented in the browser. As you see various articles and spend time on websites, a local statistic is made. At the end of the month it can be reviewed by the user (and changed by hand) and then the appropriate percentages will be sent to the websites.

  22. Re:Can we move the goalpost? on Google's AlphaGo Beats Lee Se-dol In the First Match (theverge.com) · · Score: 1

    Machine learning techniques work well with incomplete and noisy data, so they would work well even when there are occasional failures in some nodes.

  23. Re:Ads == Malware Delivery and Nuisance Content on Google, Yahoo Cry About Ad-Blocking (cnbc.com) · · Score: 1

    > what they're complaining about is that ad blockers block them just because others' ads are horrible.

    Maybe ad blockers should let through a small percentage of ads and add voting buttons on them to let users rate the ads. Those that get a good rating would be whitelisted. Then, after they collect enough data they could build an automatic classifier (like the spam filters) that predicts the user rating based on contextual features. As features they could use the name of the site where the ads are displayed, the number of times you visited that site in the last month, the actual text of the ad, its style and size and the destination link. A machine learning approach would scale up the user based rating system to ads that have not been rated yet. As people add more ratings, the system could be fine-tuned to match the mentality of its users.

  24. Re:disruption on Google, Yahoo Cry About Ad-Blocking (cnbc.com) · · Score: 1

    I always thought click based payments for ads are stupid. Instead, a pay-for-sale or pay-for-action should be used. If you have a good product, they will want to advertise it, you don't risk anything as an advertiser and you have no worries about fake clicks.

  25. Re:Punishes users and good advertisers on Google, Yahoo Cry About Ad-Blocking (cnbc.com) · · Score: 1

    > Either content creators can make money *somehow* or the content isn't generated.

    That's so false. Most content creators don't get anything other than reputation and self satisfaction from their postings. People create because creativity is one of our basic needs. We feel better when we are more creative so we create, for free or not.