Connect with us

Tech

Efficiency Spells the Difference Between Biological Neurons and Their Artificial Counterparts

Published

 on


Image by macrovector on Freepik

Machine learning has made great advances, but as this series has discussed, doesn’t have much in common with the way your brain works. Part 8 of the series explores a single facet of biological neurons which, so far, have kept them way ahead of their artificial counterparts: their efficiency.

 

Your brain contains about 86 billion neurons, which are crammed into a volume of somewhat over one liter. Although machine learning can do many things which the human brain cannot, the brain is able to perform continuous speech recognition, visual interpretation, and a host of other things, all while dissipating about 12 watts. In comparison, my laptop draws about 65 watts and my desktop machine draws over 200 watts, and neither of them is capable of running the huge ML networks which are in use today.

How does the brain achieve its remarkable efficiency? I attribute it to three essential factors:

  1. The brain is physical and chemical rather than electronic.
  2. The neurons in the brain are really slow.
  3. Neurons only require energy when they emit spikes.

While we can use electronic instruments to measure voltages within neurons, their fundamental operation is chemical. Ions are migrating from one side of a membrane to another and ionic molecules are changing orientation. This is fundamentally different from a computer, where electrons are moved about and the charge they represent travels at the speed of light. Obviously, molecules in the brain don’t require any external energy at all when there are just sitting there and the amount of energy needed to get a Sodium ion (for example) to move from one side of a membrane to another is minute.

As I mentioned in a previous article in this series, neurons spike at a maximum frequency of 250Hz and neural signals travel at a leisurely 2m/s. If we slowed our CPUs down to a similar pace, they would dissipate a lot less energy too but never as little as their biological counterparts.

The real difference, though, is that neurons need negligible energy except when they fire. Further, they don’t fire very often. By taking the total energy of the brain and dividing it by the energy needed to fire as calculated via chemistry, it can be concluded that neurons fire on average once every two seconds. It’s obvious that continuous processes like vision and hearing must be running more or less constantly using more energy. So to get things to average out, we must conclude that vast portions of the brain’s neurons seldom fire at all. Thus, a neuron that represents a specific memory (your grandmother, for example) likely fires only when you think about your grandmother.

But there’s a further way to think about this. A CPU uses some amount of energy when it is running at speed (not idle or asleep), and it uses this amount of energy regardless of the data it is processing. Adding two numbers together, adding 0+0 for example, requires essentially the same energy as adding 12,345 + 67,890. Neurons are different.

This distinction has been the genesis of the Neuromorphic computing movement. In the Brain Simulator, the processing is only required for neurons that fire, so a desktop CPU can handle up to 2.5 billion synapses per second. Neuromorphic chips capitalize on this effect to produce AI results with radically less power than conventional machine learning processes.

While neuromorphic systems have moved in the direction of more brain-like architectures, they typically are still using the ML backpropagation algorithm which is not neuromorphic at all.

“The final article in this series will summarize the many reasons why Machine Learning isn’t like your brain — along with a few similarities.”

Charles Simon is a nationally recognized entrepreneur and software developer, and the CEO of FutureAI. Simon is the author of Will the Computers Revolt?: Preparing for the Future of Artificial Intelligence, and the developer of Brain Simulator II, an AGI research software platform. For more information, visit here.

Source link

Continue Reading

Tech

New photos reveal more details about Google’s Pixel 9 Pro Fold

Published

 on

Google’s secret new line of Pixel 9 phones isn’t that big of a secret anymore. Taiwan’s National Communications Commission (NCC) released new photos of the phones including the Pixel 9 Pro Fold from almost every conceivable angle.

Android Authority found the photos in the NCC archives and uploaded galleries of each of the four phones including the Pixel 9, 9 Pro, 9 Pro XL and 9 Pro Fold. They reveal some interesting details about the new Pixel phones.

The charging rates will be a little faster than the last generation of Pixel phones: Taiwanese authorities measured 24.12W for the base model, 25.20W for the Pro and 32.67W for the 9 Pro XL. The Pixel 9 Pro Fold, however, was the slowest of all of them at 20.25W. These numbers don’t often match up perfectly with the advertised ratings, so expect Google to be promoting higher numbers at its event.

Speaking of chargers, it looks like Google needed a bigger charger to power its new phones. Photos included in the NCC leak show each phone will come with a wall charger that’s around 45W depending on which model you purchase. The charger’s plug moved from the middle to the top of the brick.

The Google Pixel 9 Pro Fold can fully unfold.
NCC/Android Authority

The latest photo dump also shows the 9 Pro Fold unfolded for the first time. Google has moved the selfie camera to the inside screen for a wider field of view. The 9 Pro Fold also has a slimmer top and bottom, a reduced fold crease on the display and a full 180 degree unfolding angle to make a screen that’s just over 250mm or just under 10 inches.

These photos are the latest in a very long list of leaks of Google Pixel 9 photos. The last Pixel 9 leak came down yesterday showing two prototype models of the base and XL models. Google might look into buying a new combination lock for the high school locker where they apparently keep all their unreleased gear.

 

728x90x4

Source link

Continue Reading

Tech

Apple Wallet now supports Canada’s Presto card, with Express Transit

Published

 on

Apple Wallet now supports the Presto transit card used in Ontario, Canada. The card can be used for travel in Toronto, Hamilton, and Ottawa.

The digital version of the card includes the Express Transit Pass feature, meaning that you can tap in and out without having to authenticate …

 

Ontario’s Presto card

The Presto contactless smart card system was first trialled back in 2007, and started the full rollout in 2009. The card can be used across 11 different transit systems in the areas covered.

Apple Wallet support was first promised many years ago, but things went quiet until a “coming soon” announcement back in May of this year.

Although the contactless terminals allow the use of credit and debit cards for regular fares, a Presto card is needed for monthly passes and discounted travel.

Apple Wallet support now available

The company made the announcement today.

Tap to ride with PRESTO on iPhone and Apple Watch.

Traveling around town just got easy with your PRESTO in Apple Wallet. With Express Mode, you don’t need to wake or unlock your iPhone or Apple Watch or open any apps to use PRESTO in Apple Wallet. Just hold your device near the reader to pay and go.

Ride, even when your iPhone needs a charge

If your iPhone needs a charge, PRESTO Card in Apple Wallet will still work. Power Reserve provides up to five hours of support, so you can still ride.

Reload on the go. 

With your PRESTO card on your iPhone and Apple Watch, you can easily load funds, right from Apple Wallet or PRESTO App. No need to visit a customer service outlet.

Extra security. Built right in 

PRESTO in Apple Wallet can take full advantage of the privacy and security features built into iPhone and Apple Watch. Your PRESTO card is stored on the device, which means Apple does not see when you use it—helping keep your data private and secure.

If you lose your iPhone or Apple Watch, you can use the Find My app to lock and help locate the device and suspend your PRESTO card or remotely erase the device and its cards.

Mobile Syrup reports that you can choose between adding your existing card to your Wallet, or creating a new one.

There are two ways to add a Presto card to Apple Wallet. You can either buy a new card or move your old one over using the Presto app.

That being said, for simplicity’s sake, unless you have a discounted Presto agreement like a student or senior plan, I think most riders will be happy just making a new card in Apple Wallet and loading funds from that app.

As with any digital card or pass, you can use either your iPhone or Apple Watch, but because each generates a unique virtual card number, you need to use the same device at both ends of your journey.

Express Transit feature

To minimize delays, Presto offers Express Transit support. This means that you don’t need to authenticate using Face ID or Touch ID on your iPhone, and you don’t need to double-tap the side button on your Apple Watch. Simply hold your device close to the pad and you’re good (a number of clues are used to detect fraudulent use).

Express Transit also has the advantage that it continues to work in Low Power mode, so you’ll still be able to complete your journey even if your phone or Watch is almost dead.

Image: Presto

728x90x4

Source link

Continue Reading

Tech

The OnePlus Pad 2 Wants to Be the iPad Air of Android Tablets

Published

 on

The original OnePlus Pad was a decent all-around Android tablet, but it was not amazing in any one area. Now, OnePlus is back with a new tablet device that packs more power, has a better screen, more speakers, and a higher starting price. OnePlus offers an Android tablet alternative that costs less than the latest iPad Airs, though it seems like it’s hewing very close to the rendition from 2023. 

The OnePlus Pad 2 is a one-size-fits-all 12.1-inch 3K tablet. At $550 for 12 GB of RAM and 256 GB of storage, it’s $70 more than the first OnePlus Pad, though it starts with more memory and twice as much internal storage as the first go around’s paltry 128 GB. It’s bigger than the 11.6 LCD on last year’s Pad, though now it’s beefed its resolution to 3K (3000 x 2120) with a stated 600 nits typical and 900 nits peak brightness. It has a variable refresh rate between 30 and 144 Hz, though it’s still an LCD screen, the same as the 2023 OnePlus Pad.

Just like last year’s version, the new Pad supports Dolby Atmos, but it boasts a six-stereo speaker configuration on either side of the device. It may not be as specifically sound-tailored as the Lenovo Tab Plus, but what’s promised is a solid middle ground. 

Last year’s tablet used MediaTek Dimensity 9000 CPU, which was good enough for most applications but not exactly top of its class. The Pad 2 is now powered with the Qualcomm Snapdragon 8 Gen 3 mobile chip. Gizmodo has already experienced some of the chip’s capabilities in Samsung’s latest foldables, and already it’s very promising. We haven’t yet had the chance to compare a Snapdragon 8 Gen 3 tablet to Apple’s latest iPad Air with M2, though on the whole, M2 usually performs better than Qualcomm’s mobile chips in bare benchmark tests. How much that matters depends on what programs you expect to use on your tablet. 

Image: OnePlus

Every device maker thinks they need AI to compete, and OnePlus isn’t an outlier here. There are promised “AI Toolbox” features like AI text-to-speech and recording summaries. The AI Eraser 2.0 will also work like Google’s Magic Eraser to remove unwanted photo elements. 

There’s a new $99 OnePlus Stylo 2 and a $150 Oneplus Smart Keyboard to accompany the new tablet. Despite the size and price difference, there will be many similarities between last year’s and the 2024 model. The Pad 2 has the same 9,510 mAh battery as last year’s, plus the 67W “SUPERVOOC” fast charging. It promises 43 days of standby time, though in our experience, the first Pad’s lifespan and promised “one-month standby life” was far more modest in practice, lasting most of the day before needing a recharge. 

With a bigger screen, the upcoming Pad 2 is slightly heavier than last year’s rendition. It weighs about 1.3 pounds, so it’s exactly between the 11- and 13-inch iPad Airs or slightly more than the base 11-inch Galaxy Tab S9 (and far less than the humongous Tab S9 Ultra). It will be relatively thin at 6.49 mm, but it’s not beating the iPad Air’s 6.1 mm or the iPad Pro 13-inch’s holy grail 5.1 mm.

The first OnePlus Pad didn’t exactly break new ground in any one category, though it did show Android tablets had legs. We’ve seen attempts from Goole and its Pixel Tablet, though that, too, wasn’t the pioneer of Android tablets. A better chip and more speakers do seem promising, though, in its effort to be everything to everyone, we’ll need to see if it manages to stand out in any area.

The OnePlus Pad 2 is now available for preorder. It should be available on the OnePlus website starting July 30 and on Amazon starting August.

 

728x90x4

Source link

Continue Reading

Trending