Profile cover photo
Profile photo
Marcelo Menegali
Marcelo's interests
View all
Marcelo's posts

Post has shared content
This is the most demonically clever computer security attack I've seen in years. It's a fabrication-time attack: that is, it's an attack which can be performed by someone who has access to the microchip fabrication facility, and it lets them insert a nearly undetectable backdoor into the chips themselves. (If you're wondering who might want to do such a thing, think "state-level actors")

The attack starts with a chip design which has already been routed -- i.e., it's gone from a high-level design in terms of registers and data, to a low-level design in terms of gates and transistors, all the way to a physical layout of how the wires and silicon will be laid out. But instead of adding a chunk of new circuitry (which would take up space), or modifying existing circuitry significantly (which could be detected), it adds nothing more than a single logic gate in a piece of empty space.

When a wire next to this booby-trap gate flips from off to on, the electromagnetic fields it emits add a little bit of charge to a capacitor inside the gate. If it just happens once, that charge bleeds off, and nothing happens. But if that wire is flipped on and off rapidly, it accumulates in the capacitor until it passes a threshold -- at which point it triggers that gate, which flips a target flip-flop (switch) inside the chip from off to on.

If you pick a wire which normally doesn't flip on and off rapidly, and you target a vulnerable switch -- say, the switch between user and supervisor mode -- then you have a modification to the chip which is too tiny to notice, which is invisible to all known forms of detection, and if you know the correct magic incantation (in software) to flip that wire rapidly, will suddenly give you supervisor-mode access to the chip. (Supervisor mode is the mode the heart of the operating system runs in; in this mode, you have access to all the computer's memory, rather than just to your own application's)

The authors of this paper came up with the idea and built an actual microchip with such a backdoor in it, using the open-source OR1200 chip as their target. I don't know if I want to guess how many three-letter agencies have already had the same idea, or what fraction of chips in the wild already have such a backdoor in them.

As +Andreas Schou said in his share, "Okay. That's it. I give up. Security is impossible."

Post has attachment

Post has shared content
Open Source Release of TensorFlow

I'm very excited to announce the open source release of the TensorFlow machine learning library that I and many others at Google have been building.

This blog post gives an overview:

The web site has a number of tutorials and documentation about the system:

The source is on GitHub:

You might also be interested in the whitepaper we've prepared that describes TensorFlow in more detail:

We'd love to hear what people think.

Post has attachment

Post has shared content
Now you can generate your own dreamy neural net image


For those of you who liked the post I shared a couple of weeks ago about the images generated by neural nets (old post:, I'm happy to announce that Alexander Mordvintsev, +Christopher Olah , and Mike Tyka have put together an open-source iPython notebook containing the code that generates these images, and you can play around with it on your own images.  (Note: this notebook depends on a few other packages, so you have to have enough persistence to install numpy and caffe to get this to work).  The iPython notebook is at, but see the blog post linked to by this post for details.

The blog post asks that people tag images they generate and share with #deepdream , so I suspect you can keep looking at that tag to see all kinds of weird and wonderful images.

Have fun, everyone!

Post has attachment
Very interesting article about mental illness and infectious diseases.

Post has shared content
Here is my recent keynote at SXSW on how to use failure as a way to drive great progress.  Full of stories about interesting failures at Google[x]!

Post has shared content
From the same author of "Manna". I'm definitely going to read it.
"The Second Intelligent Species" is my latest book on Earth's robotic future and how it will affect the human race. It is available free on my web site:

Right now, as far as we can tell, there is exactly one intelligent species in the universe and it is us: human beings. We currently see no evidence of any kind indicating that extraterrestrials exist outside of our solar system. We are the first Intelligent species.

At this moment, millions of engineers, scientists, corporations, universities and entrepreneurs are racing to create the second intelligent species right here on planet earth. And we can see the second intelligent species coming from all directions in the form of self-driving cars, automated call centers, chess-playing and Jeopardy-playing computers that beat all human players, airport kiosks, restaurant tablet systems, etc.

The frightening thing is that these robots will soon be eliminating human jobs in startling numbers. The first wave of unemployed workers is likely to be a million truck drivers who are replaced by self-driving trucks. Pilots will be eliminated soon as well. Then, as new computer vision systems come online, we will see tens of millions of workers in retail stores, fast food restaurants and construction sites replaced by robots. Unless we take steps now to change the economy, we will soon have tens of millions of workers who are unemployed and seeking welfare because they will have no other choice.

"The Second Intelligent Species: How Humans Will Become as Irrelevant as Cockroaches" explores how the future will unfold as the second intelligent species emerges and takes control. The book answers questions like:
- How will new computer vision systems affect the job market?
- How many people will become unemployed by the second intelligent species?
- What will happen to millions of newly unemployed workers?
- How can modern society and modern economies cope with run-away unemployment caused by robots?
- What will happen when the first sentient, conscious computer appears?
- What moral and ethical principles will guide the second intelligent species? 
- Why do we see no extraterrestrials in our universe?

The Second Intelligent Species offers a unique and fascinating look at the future of the human race, and the choices we will need to make to avoid massive unemployment and poverty worldwide once intelligent machines start eliminating millions of jobs.

Post has shared content
One of the best perks of being an engineer at Google is now open source. Our build system is just awesome.

Post has attachment
Wait while more posts are being loaded