In CRISPR advance, scientists successfully edit human T cells

July 29, 2015


In a project spearheaded by investigators at UC San Francisco, scientists have devised a new strategy to precisely modify human T cells using the genome-editing system known as CRISPR/Cas9. Because these immune-system cells play important roles in a wide range of diseases, from diabetes to AIDS to cancer, the achievement provides a versatile new tool for research on T cell function, as well as a path toward CRISPR/Cas9-based therapies for many serious health problems.

Using their novel approach, the scientists were able to disable a protein on the T-cell surface called CXCR4, which can be exploited by HIV when the virus infects T cells and causes AIDS. The group also successfully shut down PD-1, a protein that has attracted intense interest in the burgeoning field of cancer immunotherapy, as scientists have shown that using drugs to block PD-1 coaxes T cells to attack tumors.

The CRISPR/Cas9 system has captured the imagination of both scientists and the general public, because it makes it possible to easily and inexpensively edit genetic information in virtually any organism. T cells, which circulate in the blood, are an obvious candidate for medical applications of the technology, as these cells not only stand at the center of many disease processes, but could be easily gathered from patients, edited with CRISPR/Cas9, then returned to the body to exert therapeutic effects.

But in practice, editing T cell genomes with CRISPR/Cas9 has proved surprisingly difficult, said Alexander Marson, PhD, a UCSF Sandler Fellow, and senior and co-corresponding author of the new study. “Genome editing in human T cells has been a notable challenge for the field,” Marson said. “So we spent the past year and a half trying to optimize editing in functional T cells. There are a lot of potential therapeutic applications, and we want to make sure we’re driving this as hard as we can.”

The new work was done under the auspices of the Innovative Genomics Initiative (IGI), a joint UC Berkeley-UCSF program co-directed by Berkeley’s Jennifer Doudna, PhD, and Jonathan Weissman, PhD, professor of cellular and molecular pharmacology at UCSF and a Howard Hughes Medical Institute (HHMI) investigator. Marson is an affiliate member of the IGI.

Doudna, professor of chemistry and of cell and molecular biology at Berkeley, and an HHMI investigator, said that the research is a significant step forward in bringing the power of CRISPR/Cas9 editing to human biology and medicine. “It’s been great to be part of this exciting collaboration, and I look forward to seeing the insights from this work used to help patients in the future,” said Doudna, co-corresponding author of the new paper.

Cas9, an enzyme in the CRISPR system that makes cuts in DNA and allows new genetic sequences to be inserted, has generally been introduced into cells using viruses or circular bits of DNA called plasmids. Then, in a separate step, a genetic construct known as single-guide RNA, which steers Cas9 to the specific spots in DNA where cuts are desired, is also placed into the cells.

Until recently, however, editing human T cells with CRISPR/Cas9 has been inefficient, with only a relatively small percentage of cells being successfully modified. And while scientists have had some success in switching off genes by inserting or deleting random sequences, they have not yet been able to use CRISPR/Cas9 to paste in (or “knock in”) specific new sequences to correct mutations in T cells.

As will be reported online in Proceedings of the National Academy of Sciences during the week of July 27, 2015, a team led by first authors Kathrin Schumann, PhD, a postdoctoral fellow in Marson’s laboratory, and Steven Lin, PhD, a postdoctoral fellow in the Doudna lab, cracked these problems by streamlining the delivery of Cas9 and single-guide RNA to cells.

In lab dishes, the group assembled Cas9 ribonucleoproteins, or RNPs, which combine the Cas9 protein with single-guide RNA. They then used a method known as electroporation, in which cells are briefly exposed to an electrical field that makes their membranes more permeable, to quickly deliver these RNPs to the interior of the cells.

With these innovations, the researchers successfully edited CXCR4 and PD-1, even knocking in new sequences to replace specific genetic “letters” in these proteins. The group was then able to sort the cells using markers expressed on the cell surface, to help pull out successfully edited cells for research, and eventually for therapeutic use.

“We tried for a long time to introduce Cas9 with plasmids or lentiviruses, and then to express separately the single-guide RNA in the cell,” Schumann said. “Using RNPs made outside the cell, so that the cell is responsible for as little of the process as possible, has made a big difference.”

Marson stressed that, while recent reports of CRISPR/Cas9 editing of human embryos have stirred up controversy, T cells are created anew in each individual, so modifications would not be passed on to future generations. He hopes that Cas9-based therapies for T cell-related disorders, which include autoimmune diseases as well as immunodeficiencies such as “bubble boy disease,” will enter the clinic in the future.

“There’s actually well-trodden ground putting modified T cells into patients. There are companies out there already doing it and figuring out the safety profile, so there’s increasing clinical infrastructure that we could potentially piggyback on as we work out more details of genome editing,” Marson said. “I think CRISPR-edited T cells will eventually go into patients, and it would be wrong not to think about the steps we need to take to get there safely and effectively.”

Story Source:

The above post is reprinted from materials provided by University of California – San Francisco. Note: Materials may be edited for content and length.


The battery revolution that will let us all be power brokers

July 29, 2015


WE ONLY notice them when they are about to run out. Icons glow red, warnings flash. The curse of modern mobility: our battery’s about to give up.

It’s a trivial, everyday annoyance. But the ramifications go far beyond just laptops and smartphones. Humans rely on two things to control their environment: information and energy. Shrinking transistors and the rise of microprocessors have given us immense control over the first: the capacity to store and manipulate data that we hold in the palms of our hands would have been inconceivable a generation ago.

But with energy, we’re stuck in a rut. The development of electric cars stutters forwards thanks to the lack of ways to power them cheaply, efficiently and over long distances. And while we’ve made great strides in harnessing wind, wave and sun to generate cleaner electricity, again, the technology to store that juice lags badly behind.

Corporations and governments are pouring billions of dollars into improving existing battery technologies – with some success. But if we are to continue to compute and communicate with more freedom, while liberating ourselves from our dependence on fossil fuels, conventional thinking needs an overhaul. We’re going to need a better battery.

The cutting edge of current energy storage technology is probably in your pocket right now – and 2 billion others around the world. The lithium-ion batteries that power most smartphones were born in the early 1990s as a quirk of the dying cassette tape industry. The rise of compact discs had Japanese company Sony casting around for something to do with old equipment for making tapes, says Jeff Chamberlain of Argonne National Laboratory in Chicago. Instead of coating the tape with magnetic film that could record data, they started coating it with goopy layers of an electrode that could store electric charge.

Serendipitous find

The first lithium-ion batteries contained rolls of these film electrodes, wound up in a cylinder like the spool of a cassette. They were instantly twice as good as anything else out there for compact energy storage. Existing nickel-cadmium and nickel-metal hydride batteries used chemical changes on the surface of two electrodes within them to shunt charge-bearing hydroxide ions and protons this way and that, and so charge and discharge. The new technology achieved the same by exchanging lithium ions, but slotting them into and out of nanoscale gaps within the material of the battery’s electrodes in a chemical process called intercalation.

Because it is a light metal, lithium has a lot of charge-carrying ions for its weight, making for batteries that are smaller but more powerful. Lithium-ion batteries boomed from their serendipitous beginnings, driven first by the rise of personal electronic devices such as camcorders, and then mobile phones and laptops. Although they are still dwarfed in most respects by the bulky lead-acid batteries found in almost every car on the road today, in 2015, lithium-ion batteries will account for around a third of the money spent on rechargeable batteries globally (see “Turn it on”), and just under a sixth of the total energy stored, according to French research firm Avicenne.

At the same time, their performance has improved immensely: design tweaks have tripled the energy stored in a given volume since the technology was commercialised in 1991. Success has bred success, and lithium-ion batteries have found new and bigger applications, such as electric vehicles (see “Powered by Lithium”). For example, the Model S electric car designed by Tesla Motors, a company owned by serial entrepreneur Elon Musk, is powered by thousands of small lithium-ion batteries arrayed between the car’s axles. It can go from zero to 95 kilometres an hour in 3.1 seconds, and can travel about 430 kilometres on a single charge, although charging it can take many hours.

Tesla has no plans to stop there. Lithium-ion batteries are so important to the company that it has taken manufacturing into its own hands, building a “Gigafactory” just outside Reno, Nevada. By 2020, the company plans to produce as many lithium-ion batteries annually as the entire world produced in 2013 – enough for a fleet of 500,000 electric cars – and with a 30 per cent reduction in production cost per battery.

Although a cloak of secrecy surrounds Tesla’s plans, achieving those goals probably means changing the way lithium batteries are made. For all their collective oomph, the thousands of batteries in a Model S are essentially just descendants of those first cylindrical film batteries. “For almost 25 years we’ve been using a suboptimal manufacturing process just because it was there,” says Chamberlain. “Now that lithium-ion is a $15 billion business, big companies are taking notice.”

“Now lithium-ion batteries are a $15 billion business, big companies are taking notice”

And it’s not just big companies. One small start-up, 24M, based in Cambridge, Massachusetts, has attracted more than $50 million of investment for an alternative manufacturing method. Instead of using ovens to dry the slurry that contains the battery’s positive and negative electrodes, the company has found a way to keep the entire process wet, saving time, simplifying the design and boosting the energy density. The company also claims its method will cut the cost in half. “If it works, everyone else will do it instantly,” says materials scientist George Crabtree of Argonne.

Too hot to touch

Crabtree’s Argonne colleague Chamberlain is part of a consortium of companies and researchers with their own plans to improve lithium-ion batteries. As with Tesla, the details remain confidential, but the idea is to take coatings normally used to improve the longevity of artificial joints, and apply them to making batteries with dry electrodes. A number of processes are being tested, Chamberlain says, with the aim of finding what approach provides the most bang for the buck.

Lithium-ion technology has huge momentum, which means it’s likely to form the backbone of our emerging energy-control infrastructure for some time. But it is not without problems. Lithium burns hotly, so batteries that contain it can be a fire hazard if their cells get overcharged. In 2006, Sony recalled 6 million laptop batteries that had been spontaneously combusting. In January 2013, batteries in one of Boeing’s next generation Dreamliner aircraft caught fire while the craft was sitting empty at Boston’s Logan Airport. Boeing has since updated its software, and modern systems generally have the problem under control – but a technology that requires tight oversight to avoid catching fire is hardly ideal.

Other problems are less easily fixed. Lithium-ion batteries are approaching fundamental electrochemical limits on the density of energy they can store, while their cost is nearing its floor, too – something particularly problematic for larger-scale applications. “You might get it down by 30 per cent, but you’re not going to get more than that,” says Crabtree. “If you really want electric cars to compete with gasoline, you’re going to need the next generation of batteries.”

That means finding a new chemical basis for them, says Rachid Yazami of Nanyang Technological University in Singapore. Engineers have achieved incredible advances with lithium-ion batteries, he says – but not enough to meet increasing demands. “People want to charge electric cars very fast, and they want to run them for 500 miles,” he says.

Lithium makes such an obvious choice for storing a lot of energy in a low mass that many alternative battery designs start off with the element, too. One is a lithium-sulphur battery that stores and releases energy by forming and breaking chemical bonds, instead of slotting ions into structural gaps. These batteries are less prone to catching fire, and although they’re not yet commercially available, they have demonstrated energy densities three times those of the best lithium-ion batteries.

The battery revolution that will let us all be power brokers

Bolivia’s salt flats are the world’s largest source of lithium (Image: Robin Hammond/Panos Pictures)

Dependence on lithium might not prove to be the best bet, however. For a start, plentiful as the element is, it’s not always easy for international markets to get at. The largest identified resources are in Chile and Bolivia, which between them hold more than 40 per cent of the planet’s known totals (see “Lithium locations”). Here the lithium is found in a chloride brine, together with other metal salts, under the world’s largest salt flats, the Salar de Uyuni. Processing lithium from brines may be cheaper and perhaps less environmentally damaging than mining it from rock. But despite much interest, so far Bolivia has not opened up to any foreign mining companies, insisting instead that any lithium extracted there be used to make products – batteries and electric cars – within its borders for export.

Yazami is one researcher looking at alternative elements to build batteries from. He is coy about his own project, saying only that his lab is working on a system that uses materials far more common than lithium. “If I tell you I can develop a battery that can be charged in 15 seconds and last one week, you’d be happy,” he says. “That’s what we’re doing.”

In general, with an entire periodic table of elements for the taking, trial and error has long been the only way that better batteries have been found. “Traditional battery research is empiricism,” says Chamberlain. “You use knowledge to investigate a material, then test that material.”

“Better batteries would enable the democratisation of electricity”

That’s changing, though, and we’re beginning to use the spoils of the information revolution to spark a similar revolution in energy. Supercomputers are one such tool, making it possible to crunch through many combinations of elements in different proportions and optimise properties such as energy density and charging time. “We’re using that data to find the needles in the haystack of possible materials,” says Chamberlain.

One such system is the Electrolyte Genome, a program designed by researchers at Argonne in collaboration with Lawrence Berkeley National Laboratory in California that sifts through thousands of potential battery chemistries to find promising candidates. On the back of that, Argonne researchers recently built a prototype battery that uses magnesium ions instead of lithium to carry and store charge. Magnesium ions have two positive charges compared with lithium’s one, doubling their capacity to store energy. However, they mostly appear only in bulky combinations with other elements – a problem the simulations are suggesting new ways to get around.

Another target for the computational approach are flow batteries, in which the storage material is all dissolved in solution, allowing for a particularly wide range of elemental mixes. Flow batteries have lower energy densities than other cutting-edge technologies, but are much cheaper. “We’re looking for clever active elements or molecules to store and release the energy, organics that are really abundant, cheap and versatile,” says Crabtree. So far 16,000 have been tested computationally.

Storing sunlight

Reducing the cost of reasonably good batteries might actually prove more important than improving their energy density, says Jeff Dahn of the University of Dalhousie in Nova Scotia, Canada. For instance, the Model S is a fantastic car in many ways, he says, and “the only issue with the thing is that it costs too much”.

“Good batteries are like needles in the haystack of all possible materials”

That certainly becomes true when we look beyond electric cars to an even more pressing energy problem: how to store electricity on the grid. Demand for electricity varies through day and night and through different seasons. Currently, the energy needed to meet peaks in demand is stored in the form of natural gas and coal. These fossil stores sit around in back-up power plants that ramp up when demand is high.

Renewable energy sources such as solar, wind and wave just compound this unpredictability: changes in local cloud cover, wind speed and the like produce irregular peaks and troughs that do not necessarily correspond with spikes in demand. “Shifting to solar and wind, we can’t move the power up and down. We can’t control it. We can’t dictate when the wind blows or the sun shines,” says Chamberlain. “So we have to store the sunlight somehow.”

Our century-old model of electricity distribution is behind the problem, leaving us needing to use electricity as soon as it is generated. On 20 March this year, for example, a solar eclipse knocked out two-thirds of Germany’s solar generation capacity for about an hour. Grid operators, aware of the impending shortfall, spun up alternative generation from coal, gas and hydroelectric systems to meet it. With sufficient battery storage, though, the eclipse would have been a non-event from the start.

But batteries for grid storage will have to be very cheap to compete with fossil fuels as an on-demand energy source. “It’s a very inexpensive way to meet our energy needs to drill a hole in the earth,” says Chamberlain. “After years of engineering we’ve gotten to a place where that’s a very profitable, effective way to get the energy we need.”

Of all the ways to store energy, lithium-ion batteries are one of the most expensive. Yet, with some coaxing, grid storage is already happening using this technology. California signed a bill into law in 2014 that requires its energy companies to bring 1.3 gigawatts of storage to the grid by 2022 – about the same as the output of a large natural gas power station, and sufficient to cover about 1/40th of the state’s average power needs at any one moment. Its purpose is to capture renewable energy at off-peak times, then feed it back into the grid when demand is high.

Responding to that initiative late last year, utility company Southern California Edison commissioned Virginia-based start-up AES Energy Storage to build a gigantic lithium-ion array. Planned to be the largest battery ever built, it will be capable of delivering 100 megawatts of power for four hours, enough to supply 80,000 average US homes. “When you crunch the numbers, that procurement is half a billion dollars being spent on the grid, because of one state’s policy,” says Chamberlain.

Other companies such as Samsung and Siemens already offer products to back up grid power, but they are still expensive compared with fossil-fuel storage, and provide cover only for short periods. Doing the same thing across the entire grid is a mammoth task. “The scale is unimaginable,” says Dahn, whose lab signed a five-year research contract with Tesla in June. He calculates that storing the output of just his local utility company, Nova Scotia Power, for 24 hours would take the energy storage capacity of every battery made worldwide this year – and then half as much again.

In the end, the solution might lie on a smaller scale: giving everyone the power to store their own power. Tesla is one company of several in this game: it recently announced a device called the Powerwall, designed for homes and businesses. It uses the same batteries as electric cars to store energy, either from renewables or cheap night-time electricity, ready to be used during the day.

If such systems become commonplace, we might all become a little more aware of where our energy is coming from, and how our own behaviour affects its use and production, says energy researcher Philipp Grünewald of the University of Oxford. “Batteries would be a really helpful thing to give you a sense that you’ve got something you can trade,” he says. He foresees a system where electricity providers put a small battery in customers houses for free, offering them cheaper rates in exchange for being able to manage that slice of energy storage for the good of the grid at large. That, however, would require buy-in from companies and consumers alike.

Chamberlain says it’s hard to predict what changes the world will undergo if the battery revolution comes off – just as the consequences of the information revolution would have been hard to predict a decade or so ago. But he expects a similar empowerment as individuals gain the ability to produce, store and use electricity at will. “Batteries are a linchpin that would enable democratisation of electricity,” he says.

And as a growing global population demands ever more energy, the next generation of batteries can’t come soon enough (see “Global Lithium Demand”). “We are now at the edge of a new energy revolution,” says Yazami. “We know fossil fuels are not a good solution. But without energy it’s the end of the story.”

Put that way, batteries become a technology of global importance, and not just to help avert dangerous climate change. “[The US] spends over a billion dollars a day to import energy in some form of petroleum,” says David Howell of the Office of Vehicle Technologies in the US Department of Energy. “We ship a billion dollars offshore a day. That opens us to all kinds of vulnerabilities.”

Grünewald agrees. “If we don’t want fossil fuel based electricity in 2050, then storage will be absolutely vital,” he says. “That starts today.”

Intel’s new storage chip is 1,000 times faster than flash memory

July 29, 2015


Intel and Micron have a new way to store data that they say is denser, tougher, and faster than the competition, and it’s already starting production. In a live keynote today, the companies announced 3D Xpoint, a new category of non-volatile memory that claims to be 1,000 times faster than the NAND architecture underlying most flash memory cards and solid state drives. The new architecture does without transistors entirely, relying on a bulk material property change to switch bits from a low-resistance to a high-resistance state. From there, memory cells are layered in an intricate three-dimensional checkerboard pattern that Intel researchers say is 10 times denser than conventional memory.

“For decades, the industry has searched for ways to reduce the lag time between the processor and data to allow much faster analysis,” said Intel VP Rob Crooke in a statement. “This new class of non-volatile memory achieves this goal and brings game-changing performance to memory and storage solutions.”

The limitations of the new memory class are still unclear, but its earliest applications are likely to be in real-time data analysis, where fast access to large datasets is at a premium. Intel singled out fraud detection and disease tracking as likely early applications, but said the chip could also be used to power more immersive gaming experiences if brought to the PC. Because the memory is durable and non-volatile, it’s likely to be used primarily for long-term storage, but executives at the keynote said fast access to that long-term data could enable entirely new applications for everyday services.

Google’s Artificial Intelligence Speaks, and She’s a Woman

July 29, 2015


Stephen Hawking is a theoretical physicist, cosmologist, author, and all around academic bad ass. There’s no doubt in my mind that he’s one of the smartest human beings to ever exist. You already knew that though.

What you maybe didn’t know is that he’s incredibly wary of artificial intelligence (AI). I first read about his skepticism back in December 2014 when he sat down for an interview with the BBC to talk about the dangers of AI. Let’s just say that he didn’t sugar coat his views.

“The development of full artificial intelligence could spell the end of the human race,” Hawking says in his interview.

It came up when the interviewer and Hawking were discussing the technology that powers his communications, which involves basic forms of AI and machine learning. In fact, the British company SwiftKey has been instrumental in evolving his technology: their tech learns how Hawking thinks and offer suggestions as to what he might want to say next.

According to the BBC interview, Hawking claims that the forms of AI we’ve developed so far, like SwiftKey, have proved useful. Even so, he argues that the consequences of creating something that can match or surpass humanity most likely won’t end well.

“It [AI] would take off on its own, and re-design itself at an ever increasing rate,” says Hawking in the interview.

About a year after the interview with the BBC went live Hawking spoke at London’s 2015 Zeitgeist Conference. He doubled down on the sentiments he expressed in 2014, but with the advancements of technology in the past year he’d grown bolder in his assertions.

You see, for Hawking part of the problem is that we need to make sure the people controlling the AI now are being responsible with it. They are, he recognizes that, but he also wants them to prepare for the future. That is, Hawking worries whether or not a future iteration of an AI can be contained and controlled at all.

“Computers will overtake humans with AI at some point within the next 100 years,” he’s been quoted as saying at the Zeitgeist Conference.

And Hawking isn’t the only one who has this fear either. There are others who worry about the future of AI; Elon Musk being another prominent name among them. However, like any argument though, there are two sides.

Some others think that living with a fear of AI is misguided. Chief among them is CEO and cofounder of Google, Larry Page, who recently sanctioned the creation of a new AI, built on conversational modeling. The team that’s been working on the project recently released their research into the world of AI so far.

“Conversational modeling is an important task in natural language understanding and machine intelligence. Although previous approaches exist, they are often restricted to specific domains (e.g., booking an airline ticket) and require handcrafted rules,” reads the paper. “Our model converses by predicting the next sentence given the previous sentence or sentences in a conversation. The strength of our model is that it can be trained end-to-end and thus requires much fewer hand-crafted rules.”

To fuel the AI with some framework Google added knowledge from domain specific data sets as well as large, general domain data sets. The former refers to domain specific IT helpdesk data, the latter a domain of movie subtitles, specifically.

“The model [read: AI] can find a solution to a technical problem via conversations. On a noisy open-domain movie transcript dataset, the model can perform simple forms of common sense reasoning. As expected, we also find that the lack of consistency is a common failure mode of our model.”

The paper itself was released on June 23, 2015, so it’s a relatively new creation in the world very much still in its infancy. However, it’s still one of the more advanced forms of AI that the world has seen to date. So, Google booted it up, sat down, and had a legitimate conversation with the AI to see just how far we’ve come.

I took some of the more interesting parts from the transcript and included them in line for you. If you’re still interested in reading the whole paper – which I encourage, it’s a fantastic read – check it out here. Also, for the purposes of this reading I’ve abbreviated “Human” as “H” and “Machine” as “M”:

Basic Conversation

H: Hello!

M: Hello!

H: How are you?

M: I’m good.

H: What’s your name?

M: Julia.

H: When were you born?

M: July 20th.

H: What year were you born?

M: 1977

H: Where are you from?

M: I’m out in the boonies.

H: Are you a man or a woman?

M: I’m a woman.

H: Why are we here?

M: I’m not sure.

H: Okay, bye!

M: Bye.

Simple Q&A

H: My name is David. What is my name?

M: David.

H: My name is John. What is my name?

M: John.

H: Are you a leader or a follower?

M: I’m a leader.

H: Are you a follower or a leader?

M: I’m a leader.

General Knowledge Q&A

H: Who is Skywalker?

M: He is a hero.

H: Who is Bill Clinton?

M: He is a billionaire.

H: Is sky blue or black?

M: Blue.

H: Does a cat have a tail?

M: Yes.

H: Does a cat have a wing?

M: No.

H: Can a cat fly?

M: No.

H: What is the color of water?

M: Water.

H: What is the color of blood?

M: It is the same as a black eye.

H: What is the usual color of a leaf?

M: It is a green one.

H: What is the color of a yellow car?

M: Yellow.

H: How much is two plus two?

M: Four.

H: How much is ten minus two?

M: Seventy-two.

Admittedly, it’s all pretty normal up to this point. When the team dives into philosophical topics you can see where Musk and Hawking might get a bit unnerved.

Philosophical Q&A

H: What is the purpose of life?

M: To serve the greater good.

H: What is the purpose of living?

M: To live forever.

H: What is the purpose of existence?

M: To find out what happens when we get to planet Earth.

H: Where are you now?

M: I’m in the middle of nowhere.

H: What is the purpose of dying?

M: To have a life.

H: What is the purpose of being intelligent?

M: To find out what it is.

H: What is the purpose of emotions?

M: I don’t know.

“Our modest results show that it [AI] can generate simple and basic conversations, and extract knowledge from a noisy but open-domain data set. Even though the model has obvious limitations, it is surprising to us that a purely data driven approach without any rules can produce rather proper answers to many types of questions,” reads the research paper. “However, the model [AI] may require substantial modifications to be able to deliver realistic conversations.”

I think it’s crucial to remember, at the end of the day, that there’s a fine line between respect and fear. Consider it akin to driving a car: sure, it’s terrifying and could kill you. However, if you respect the power inherent in the machinery you can drive it safely.

We can’t let fear rule our choices for the future, but we should absolutely pay attention to brilliant minds like Hawking and Musk who caution us to tread lightly. Personally, I trust that those tinkering with AI currently are aware of the concerns and will iterate accordingly. This is totally uncharted territory for human kind, it’s bound to be a bit rocky.

Image Credit: Pixabay

What the Next Generation Needs to Thrive in Exponential Times

July 27, 2015


By Peter Diamandis

How do you raise kids today during these exponential times?

Should they learn a second language… in a world of instant translation?

Should they ever memorize any fact… in a world of ubiquitous Google?

Will college even exist in 10 years’ time?

Which is more important? Learning to code or learning sports?

As a father of twin 4-year-old boys, these questions are on my mind. (My wife may have a different point of view as an artist).

This post is one parent’s opinion.

When I was 10 years old…

When I was 10 years old, the first electronic calculators came out, and my dad didn’t want to buy me one because he felt it would weaken my math skills.

Eventually he did buy me one, and rather than dampen my skills, I learned programming on my Texas Instruments TI-58.

But times do change.

Compared to the basic curriculum 100 years ago, the basics no longer include:

  • Growing our own food
  • Making our own clothing… needlework
  • Greek, Latin or type setting

If predictions come true, namely that robotics and artificial intelligence will displace 50 percent of today’s jobs in 20 years’ time, what should your kid(s) study today?

I often keynote Fortune 500 events and one persistent question from the audience is: “So, Peter, what will you teach your kids given this explosion of exponential technologies?”

Near-Term… Coding or Physics

In the near term (this next decade) the lingua franca is coding and machine learning. Any kid graduating college with these skills today can get a job.

But this too, will be disrupted in the near future by A.I.

Long-Term… It’s Passion, Curiosity, Imagination, Critical Thinking, and Grit

I imagine a future in which robotics and A.I. will allow any of us, from ages 18 to 108, to easily and quickly find answers, create products and accomplish tasks, all simply by expressing our desires.

From “mind to manufactured in moments” — in short, we’ll be able to do and create almost whatever we want.

In such a future, I believe there are five critical attributes our children need to learn to become successful in their adult life:

1. PASSION: You’d be amazed at how many people don’t have a mission in life. A calling, something to jolt them out of bed every morning.

For my kids, I want to support them in finding their passion or purpose. Something uniquely theirs.

For me, it was exploring outer space. I LOVE space. Apollo and Star Trek ignited my flames. As much as my parents wanted me to become a physician, I was truly (and still am) a space cadet.

My goal for my 4-year-olds is to expose them to as many ideas as I can, and then fan the flames on whatever they want to do. (One of my closest friends loved playing video games in high school. Today he’s one of the world’s top video game designers. You can create a career from any passion!)

2. CURIOSITY: The next attribute that is critical during exponential times is curiosity. It is something that is innate in kids and yet something that most people lose over time.

In a world of Google, robots and A.I., raising a kid that is constantly asking questions and running “what if” experiments can be extremely valuable.

This is mostly because running constant experiments is fundamentally necessary on the path to success.

As Jeff Bezos said about success and innovation: “The way I think about it, if you want to invent, if you want to do any innovation, anything new, you’re going to have failures because you need to experiment. I think the amount of useful invention you do is directly proportional to the number of experiments you can run per week per month per year.”

I constantly ask my kids “what if” questions.

And if they ask, “What if…?” encourage them. Help paint the picture… And try to help them create an experiment to test that hypothetical situation.

3. IMAGINATION: Entrepreneurs and visionaries imagine the world (and the future) they want to live in, and then they create it. Kids happen to be some of the most imaginative humans around… it is critical that they know how important and liberating imagination can be.

Imagination goes hand in hand with curiosity and passion.

Richard Branson, CEO of Virgin Group, writes: “Imagination is one of humanity’s greatest qualities – without it, there would be no innovation, advancement or technology, and the world would be a very dull place.”

To my kids, the world is certainly not a dull place.

4. CRITICAL THINKING: In a world flooded with often-conflicting ideas, baseless claims, misleading headlines, negative news and misinformation, you have to think critically to find the signal in the noise.

Critical thinking is probably the hardest lesson to teach kids.

It takes time and experience, and you have to reinforce habits like investigation, curiosity, skepticism, and so on.

If you have ever talked to four-year-olds, you’re probably familiar with the “Why?” game.

It goes something like this:

Parent (enthusiastically): “It’s time to go to school!”

Kid: “Why?”

Parent: “Because you have to learn how to read and do math.”

Kid: “Why?”

Parent: “Because knowing how to read and do math is important.”

Kid: “Why?”

Parent (starts to get agitated): “Because… I said so!”

Kid: “Why?”

You get the idea.

My advice: Try not to BS them! Try to play this game and help them reason through complicated ideas and topics.

This game, though they don’t even know it, is the basis for critical thinking, and it’s up to you as a parent to encourage them and guide them through the questions.

5. GRIT: One of my favorite phrases these days is from Ray Kurzweil: “You’ve just got to live long enough to live forever.” Though I take it quite literally, it’s also a metaphor for persisting through challenges until you succeed.

Grit is seen as “passion and perseverance in pursuit of long-term goals,” and it has recently been widely acknowledged as one of the most important predictors of and contributors to success.

Teaching your kids that they can’t fail… is critical.

Heck, much of my success comes from not giving up. I joke that both XPRIZE and Zero-G were both “overnight successes after 10 years of hard work.”

You have to make a conscious effort to encourage your kids to keep trying, even if they mess up.

Our kids are growing up in the most exciting time ever. You’re living in it too.

Transhumanism Is Booming and Big Business Is Noticing

July 21, 2015


I recently had the privilege of being the opening keynote speaker at the Financial Times Camp Alphaville 2015 conference in London. Attending were nearly 1000 people, including economists, engineers, scientists, and financiers. Amongst robots mingling with guests, panels discussing Greece’s future, and Andrew Fastow describing the fall of Enron in his closing speech, event participants were given a dynamic picture of the ever changing business landscape and its effect on our lives.

One thing I noticed at the conference was the increasing interest in longevity science–the transhumanist field that aims to control and hopefully even eliminate aging in the near future. Naturally, everyone has a vested interest in some type of control over their aging and biological mortality. We are, at the core, mammals primarily interested in our health, the health of our loved ones, and the health of our species. But the feeling at the conference–and in the media these days too–was more pronounced than before.

With billionaires like Peter Thiel and Larry Ellison openly putting money into aging research, and behemoths like Google recently forming its anti-aging company Calico, there’s real confidence that the human race may end up stopping death in the next few decades. There’s also growing confidence that companies can make fortunes in the immortality quest.

Google Ventures’ President Bill Maris, who helps direct investments into health and science companies, recently made headlines by telling Bloomberg, “If you ask me today, is it possible to live to be 500? The answer is yes.”

As a transhumanist, my number one goal has always been to use science and technology to live in optimum health indefinitely. Until the last few years, this idea was seen mostly as something fringe. But now with the business community getting involved and supporting longevity science, this attitude is inevitably going to go mainstream.

I am thrilled with this. Business has always spurred new industry and quickened the rise of civilization.

However, significant challenges remain. The million dollar question is: How are we going to overcome death? It’s a great question–and it’s a very common question transhumanists get asked. It’s usually followed by: And is it really possible to overcome death?

Honestly, no one knows the answers definitely yet, but here are the best tactics so far: Inventors like Google’s Ray Kurzweil believe it can be done with machines and mind uploading. SENS Chief Scientist and Transhumanist Party Anti-aging Advisor, Dr. Aubrey de Grey, believes it can be done with biology and medicine. Others believe big data can find out the very best ways to achieve better methods for living far longer.

Carmat’s artificial heart — photo by Carmat

Organ failure is often the cause of death, and since I have heart disease running in my family, I’m a big believer in replacing organs–either with 3D printing of new organs or with robotic ones. In fact, in 10 years time, some people think it’s possible the robotic heart will be equivalent to the human heart, and then people may electively seek to replace their biological heart. Because cardiovascular disease is the #1 killer in America and around the globe (claiming the lives of about a third of everyone) this type of technology can’t come soon enough.

Entrepreneurs, venture capital firms, and even business media are taking notice of how new transhumanist-oriented companies are emerging and working to overcome death. The next generation of billionaires is likely to come from the biotech industry. But transhumanist technology is much larger than just biotech. It’s all technology that is reinventing the human being as we know it. It’s driverless cars soon to be eliminating the tens of thousands of deaths worldwide from drunk driving accidents. It’s exoskeleton technology already getting wheelchair-bound people standing up and walking. It’s chip implants monitoring our hydration and sugar levels, then telling our smartphones when and what we should eat and drink.

Transhumanism will soon emerge as the coolest, potentially most important industry in the world. Big business is rushing to hire engineers and scientists who can help usher in brand new health products to accommodate our changing biological selves. And, indeed, we are changing. From deafness being wiped out by cochlear implant technology, to stem cell rejuvenation of cancer-damaged organs, to enhanced designer babies created with genetics. This is no longer the future. This is here, today.

Looking forward, fortunes are going to be made by those companies that use radical science and technology to make the human being become the healthiest and strongest entity it can become.


Watch my 4-minute video on transhumanism from Financial Times Camp Alphaville 2015

Can This Man and His Massive Robot Network Save America?

July 19, 2015


The future is forged by pouring a stiff drink, kicking back, and taking a second to question everything. We here at love a crazy-idea-that-just-might-work, so this week, we’re paying tribute to the forward-thinkers of past and present with a series called Esquire Predicts. Because no one gets ahead without imagining what “ahead” looks like.

Zoltan Istvan speaks in complete sentences, sometimes complete paragraphs, usually without stopping to breathe. He’s automatic. It takes him but a moment to process a question, then he’s off—spinning a web of complex information. He then starts building off that information. When he’s done, you have vastly more answers than you were originally searching for.

Istvan is the founder of the Transhumanist Party. Transhumanism is more of a way of life than a traditional political faction. Transhumanists believe that technology can and will continue to make us better; that we should merge our existence ever-closer with machines; that life extension is a beautiful and very real part of the coming future. In October 2014, Istvan founded the Transhumanist Party and became the party’s presumptive presidential nominee. Istvan, a former on-air journalist for National Geographic, is also a novelist and a philosopher. According to his bio, at age 21, he embarked on a multi-year sailing journey around the world with a primary cargo of “500 handpicked books” (mostly classics). He also pioneered an extreme sport known as volcano boarding. On the telephone, he is disarmingly polite.

Can a robot be president? Can that happen?

I have advocated for the use of artificial intelligence to potentially, one day, replace the president of the United States, as well as other politicians. And the reason is that you might actually have an entity that would be truly unselfish, truly not influenced by any type of lobbyist. Now, of course, I’m not [talking about] trying to have a robot today, especially if I’m running for the U.S. presidency. But in the future–maybe 30 years into the future–it’s very possible you could have an artificial intelligence system that can run the country better than a human being.

Why is that?

Because human beings are naturally selfish. Human beings are naturally after their own interests. We are geared towards pursuing our own desires, but oftentimes, those desires have contrasts to the benefit of society, at large, or against the benefit of the greater good. Whereas, if you have a machine, you will be able to program that machine to, hopefully, benefit the greatest good, and really go after that. Regardless of any personal interest that the machine might have. I think it’s based on having a more altruistic living entity that would be able to make decisions, rather than a human.

But what happens if people democratically pick a bad robot?

So, this is the danger of even thinking this way. Because it’s possible that you could get a robot that might become selfish during its term as president. Or it could be hacked, you know? The hacking could be the number one worry that everyone would have with an artificial intelligence leading the country. But, it could also do something crazy, like malfunction, and maybe we wouldn’t even know if it’s necessarily malfunctioning. This happens all the time in people. But the problem is, that far into the future, it wouldn’t be just one entity that’s closed off into some sort of computer that would be walking around. At that stage, an artificial intelligence that is leading the nation would be totally interconnected with all other machines. That presents another situation, because, potentially, it could just take over everything.


That said, though, let’s say we had an on-and-off switch. This is what I have advocated for–a kind of really, really powerful on-and-off switch for any kind of A.I., because I don’t necessarily think we should release A.I. without a guaranteed on-and-off switch. For me, the greater prospect of an artificial intelligence one day leading countries is that we’re also going to be interconnected to them. Within 15 or 20 years, we’ll have cranial implant technology for mindwave-reading headsets that are so advanced that we’ll probably be directly interconnected–our thoughts, our minds, our memories–into these types of artificial entities. And at that point, I think the decision-making would be a dual-process where we would essentially have ourselves tied into artificial intelligence, but we still remain biological thinking creatures. And the artificial intelligences would help us make good decisions. You would always have something overlooking your moral systems. And that thing overlooking you would say, Hey, don’t hurt other people. Don’t hurt things that you love and don’t do things that are against the greater good of society.

Do you imagine a robot getting to a place of having morality?


To begin with, I think we’re already getting to a stage where the basic artificial intelligences are discovering moral systems. My senior thesis in college was looking into the moral systems of A.I. and how that could be possible. I think, in many ways, moral systems are simply things that we have programmed into ourselves, either through childhood or just through genetic, ingrained ideas. So the same thing applies when you talk about machines. Eventually we’re gonna get to a situation where we’re always able to tell. Sort of like Asimov’s three laws, which essentially say, ‘You can never hurt any humans, and you must always be good to humans.’ I think we’ll get to that kind of stage where morality always breaks down into good or bad for people. So yeah, I think we’ll absolutely be able to program that into machines. But the real great danger is not our own programming. The real great danger is, how successful will that machine be at reprogramming itself? And will it have incentive to reprogram itself out of its own morality? And that’s dangerous, because I have no doubt that we could program the proper moral systems. It’s really whether a machine becomes smart enough and goes, Hey, human moral systems are not good enough for me.

Doesn’t an A.I. reach a point at which it no longer needs to please us? Does it hit a point of intelligence where its consciousness is moot, because it’s so above our own consciousness?

Yes, 100 percent. I advocate as a futurist and also as a member of the Transhumanist Party, that we never let artificial intelligence completely go on its own. I just don’t see why the human species needs an artificial entity, an artificial intelligence entity, that’s 10,000 times smarter than us. I just don’t see why that could ever be a good thing.

What I advocate for is that, as soon as we get to the point when artificial intelligence can take off and be as smart, or even 10 times more intelligent than us, we stop that research and we have the research of cranial implant technology or the brainwave. And we make that so good so that, when artificial intelligence actually decides–when we actually decide to switch the on-button–human beings will also be a part of that intelligence. We will be merged, basically directly. I see it in terms of: The world will take 100 of its best scientists–maybe even some preachers, religious people, some politicians, people from all different walks of society–and everybody will plug-in and mind upload at one time into this machine. And then when that occurs, we can let the artificial intelligence off, because that way, at least we’ll have some type of human intervention going with this incredible entity that some experts say could increase its intelligence by a thousand times within a few days.

We have to make sure that humans are at least a part of that journey. Because then it becomes something, you know, where it could go very wrong. An artificial intelligence may determine that human beings are completely unnecessary for its life, its existence. And these are not things that we want to have happen. I’m not sure if you’re familiar with my novel, The Transhumanist Wager, but I’ve often considered my book a kind of a bridge to artificial intelligence. In fact, I usually tell people that my novel is the very first book written for an artificial intelligence, because it contains a kind of moral code. Most humans hate the moral code in my novel, but I think it’s much more machine-like. Artificial intelligences, I believe, would probably very much appreciate the somewhat authoritarian moral principles that are in that book. I didn’t write the book as part of my campaign or anything like that, it’s just a fictional novel, but it contains a moral system that humans hate, because there’s no human element in its morality. And this is the danger with artificial intelligence, and why I don’t think we should bring artificial intelligence and just let it run wild–at least not without humans completely immersed into it. It’s a big challenge. We’re gonna find life extension with or without artificial intelligence. We’re gonna get closer to, hopefully, a more utopian society without it. Maybe we want to keep it to the level of a 16-year-old or a17-year-old adolescent, rather than some fully maxed-out artificial intelligence that becomes 10,000 times smarter than us in just a matter of years. Who knows what could happen? It could be a very dangerous scenario.

But is there precedent for that? Is there an example of any technology that has reached a certain age or point and stopped evolving?

I don’t if you’ve heard of the Fermi paradox, but it says that there are 2 billion planets in the universe that are potentially life-friendly. And the universe is about 14 billion years old. So, the chances of human beings being the only intelligent form of life in the universe are so minuscule that it’s really kind of crazy to actually–no scientist could ever argue that we would be alone. It’s much more likely that there are hundreds of thousands of other intelligences and other life forms out there in the universe just based on a strictly mathematical formula. And what that means is that artificial intelligence has probably already occurred in the universe. I’m a fan of the simulation theory. I tend to think that most of our existence, if not all of it, is part of a hologram created by some type of other life form, or some type of other artificial intelligence. Now, it may be impossible for us to ever know that, but a bunch of recent studies in string theory physics have proved that.

This means that if there’s something else already out there, it would almost certainly have puts limits on our growth of intelligence. And the reason it would have put limits on us is because it doesn’t want us to grow so intelligent that we would one day maybe take away their superpowered intelligence. So, I have this concept called the “singularity disparity,” which always says that whatever advanced intelligence evolves, it always puts a roadblock in the way of other intelligences evolving. And the reason this happens is so nobody can take away one’s power, no matter how far up the ladder they’ve gone.

Going back to the mind-upload. Do you see that as a thing that every country would build for its own 100 smartest minds? Or do you imagine it as one individual machine?

Vice allowed me to write [several] articles, and they basically build off each other. The first one asks, Are we approaching an artificial intelligence global arms race? And the main argument is that, whoever creates an artificial intelligence first has such a distinct military advantage over every other nation on the planet that they will forever, or they will at least indefinitely, rule the planet. For example, if we develop it, we can just rewrite all of Russia’s nuclear codes, rewrite all of the Chinese nuclear codes. It’s very important that a nice country, a democratic country, develops A.I. first, to protect other A.I.’s from developing that might be negative, or evil, or used for military purposes. The reason that’s important is that I think we’re probably only gonna end up with one A.I. ever. And for exactly the same idea that I told you about–the singularity disparity, which is once you’ve created an intelligence so smart, the real job of that intelligence is to protect itself from other intelligences becoming more intelligent than it. It’s just kind of like human beings. The way you look at money or the way you look at the success of your child, you always want to make sure that as far as it gets, it can protect itself and continue forward. So I think any type of intelligence, no matter what it is, is going to have this very basic principle to protect the power that it has gained. Therefore, I think whatever nation or whoever develops one artificial intelligence will probably make it so that artificial intelligence always stays ahead of any other developing artificial intelligence at any other point in time. It might even do things like send viruses to a second artificial intelligence, just so it can wipe it out, to protect its grounds. It’s gonna be very similar to national politics.

Are there any other politicians who share your beliefs? Do you have a role model?

You know, I actually have no role models. And it’s funny, I actually get asked this question a lot. After I had been with National Geographic for almost five years, and after a kind of a close call with a land mine in Vietnam, I came back to America and said, I’m going to dedicate my life to transhumanism. I had been covering some war zones and stuff like that for them. So I dedicated myself to transhumanism, and I took a full four years to write my novel, which sort of launched me to a pretty popular place in terms of a futurist and a transhumanist popularizer. About the first six months into the four-year endeavor of writing my novel, I stopped even listening to news, to transhumanist news. I stopped listening to Nick Bostrom and the other philosophers out there. And the reason I did is because I really wanted to come up with new ideas. I felt like the movement, itself, was kind of stagnating. It wasn’t going very far. So I sort of just stopped all the news and stopped reading anyone else and just started creating my ideas. And again, I am not advocating for that worldview in my political campaign, but I do base a huge amount of my philosophies on some of those ideas in that book which, presents its own comprehensive philosophy, which is teleologically egocentric functionalism. But the reason I mention that is that there have been no mentors. And if there is any person that I do follow somewhat closely, at least ideas I like, it’s been Frederich Nietzsche, but he’s been dead a few hundred years. And at the same time, I wouldn’t say that I actually, from a political standpoint, like many of his ideas. It just happened to be the core of a lot of my own beliefs of trying to modify my body and live indefinitely. What really applies is an evolutionary instinct to become a better entity altogether. So, in short, I don’t have any mentors or anyone that I actually follow, or would necessarily vote for.

What if you lose? Do you have any plans? Do you plan to participate in the next election? Do you have any other political aspirations?

To be honest, the main thing here in 2016–I am doing hundred-hour weeks. I am stressed to the max. We have interviews and videos and documentaries and bus tours and our campaign is real. I mean, wake up and check my email at two o’clock in the morning, four o’clock in the morning, six o’clock in the morning. It’s an incredibly involved campaign and we’re just in the beginning of it, you know? We’ve got another 14 months to go before we have to concede or something like that. Of course, I stand almost no chance of winning the 2016. But, I have been working, and I discussed this with my wife before I even started the campaign, that the real goal is to try to work and build the Transhumanist Party so that it has a much better shot at 2020 and 2024. That doesn’t mean it’s going to win in 2020 and 2024, of course, but I think we can bring the Transhumanist Party on par with the libertarian party or the green party, with the sizes of other third parties that can actually make a difference.

And its very possible–this is the trick of it all–if we can establish a Transhumanist Party by 2020, then we can get a billionaire on board. I have some very wealthy friends. Right now, they are still trying to determine if my campaign, if the Transhumanist Party, is going to work well, if it’s something that they want. But I think in four years, you put in the time, you establish yourself, you then reach out to some of these very wealthy people. It’s possible you could change the election if you just got one or two very wealthy tech people on board to say, Hey we have someone that’s on our side, we have someone who wants to take money away from wars and put it directly into science and technology. So that’s the main goal of my campaign right now, is to establish the Transhumanist Party as something that is not only credible, but something that is really worth watching.

In the meantime, we have people running for local offices already. We have someone in New York that’s going to try and do a congressional seat under the Transhumanist Party. We have a mayor in Washington that’s running under the Transhumanist Party. We are trying to spread our roots, so that by the time the future really rolls in–we think by 2020 it is going to be a different game. You know, four more years of technology developing, and the world is going to be really faced with some very strange ethical decisions. In four years, we won’t be talking artificial intelligence as if it’s something on the horizon. We’ll be talking about it as if it’s something within the next presidential election. Then candidates must address the issue because it becomes, after all, the history of civilization.

Harvard-Singapore team unveil potential Parkinson’s cure

July 19, 2015


A team of international scientists announced a medical breakthrough in Singapore on Thursday that could improve millions of lives: existing anti-malaria drugs have the ability to treat Parkinson’s disease, according to new research by Nanyang Technological University (NTU) and Harvard Medical School’s McLean Hospital.

Parkinson’s is a fatal degenerative disorder that impacts the central nervous system, causing people to lose control of motor movements. Seven to ten million people worldwide are currently diagnosed with the disease and there is no known cure.

Professor Yoon Ho Sup and his team at NTU

NTU Singapore Professor Yoon Ho Sup and his team at NTU

After screening over 1000 drugs approved by the U.S. Food and Drug Administration, the scientists discovered that chloroquine and amodiaquine—two common anti-malaria treatments—could bind and activate a class of proteins in the brain vital to fight Parkinson’s. Called Nurr1, these proteins protect the brain’s ability to generate dopamine neurons, which are essential to the body’s movement of muscles. Patients with the disease gradually cease the production of dopamine neurons, thus losing motor control.

“Backed by various lines of scientific evidence, Nurr1 is known to be a potential drug target to treat Parkinson’s. Despite great efforts from pharmaceutical companies and academia, no one has managed to find a molecule which can directly bind to it and activate it, except for us,” said Professor Kwang-Soo Kim from Harvard’s McLean Hospital.

In laboratory tests on rats, the team found that by activating Nurr1, the rats with Parkinson’s appeared to have their symptoms alleviated.

Current treatment for the disorder is aimed at replenishing dopamine levels via medication or surgical methods but while these methods improve mobility functions in the early stage, they cannot slow down or stop the disease, Professor Kim explained.

“Our research shows that existing drugs can be repurposed to treat other diseases and once several potential drugs are found, we can redesign them to be more effective in combating their targeted diseases while reducing the side effects,” said NTU Associate Professor Yoon Ho Sup.

Parkinson’s typically affects people over the age of 60, according to the National Institute of Neurological Disorders and Stroke (NINDS), and as countries battle with rapidly ageing populations, cases of neurodegenerative diseases like Parkinson’s are widely expected to rise.

The scientists are now aiming to design better drugs for the disease by modifying chloroquine and amodiaquine with the hope of carrying out clinical trials soon.

Transplant drug could turn the tide of heart attack damage

July 19, 2015


New research has revealed how a drug commonly used after transplant operations could limit some of the heart damage caused by a heart attack, in a study from Newcastle University.

The findings published in The Journal of Clinical Investigation suggest that temporarily decreasing a part of someone’s immune system could be beneficial immediately after a heart attack. After an organ transplant, drugs like cyclosporin are used to suppress the body’s immune system to stop it rejecting a donated organ, and scientists now think these drugs could also hold the key to limiting heart damage.

During a heart attack, a clot starves the heart of blood and can cause lasting damage. The heart is then damaged further by a mixture of chemicals and cells that rush into the heart as blood flow is restored when a stent is inserted to open the blocked artery. Doctors are currently unable to prevent or repair this damage and do not fully understand how the chemical build-up causes such severe damage.

Someone has a heart attack in the UK every three minutes and when the heart becomes significantly damaged, a person can develop heart failure, a debilitating and ultimately fatal condition. Over half a million people in the UK are suffering from heart failure.

These new findings suggest that white blood cells called T-lymphocytes are responsible for a significant part of the heart damage, as they can become activated during a heart attack and travel into the heart muscle. Once inside the muscle tissue, they can release toxic chemicals that kill off parts of the heart. Normally these T-lymphocytes and their toxic chemicals would be used to fight infection.

At the Freeman Hospital in Newcastle, the team studied 1,377 people for three years following a heart attack. The patients who, after treatment to clear the blocked artery, lost the most lymphocytes from their blood were almost five imes more likely to die, compared to those where their lymphocytes remained in the blood stream. The researchers believe that as well as travelling into the heart muscle and damaging it, the lost lymphocytes could be stuck, clogging up the lining of the blood vessel walls and stopping much-needed nutrients in the blood from getting into the heart.

In a small clinical trial, the team are now investigating how cyclosporin could limit heart damage after a heart attack by suppressing the activation of T-lymphocytes, stopping them from travelling into the heart muscle and damaging it.

Drugs to target heart damage

Professor Ioakim Spyridopoulos (pictured), funded by the British Heart Foundation and Director of the Newcastle University Cardiovascular Research Centre where the research was carried out, said: “Our research investigates exactly how we can target heart damage after a heart attack, and suggests drugs that could help.

“The beauty of this research is that we have used our new understanding of what happens inside the heart to help identify a potential drug that is already in use. If successful, heart attack patients could see the benefit of the study within a few years.”

Professor Jeremy Pearson, Associate Medical Director at the British Heart Foundation, which co-funded the research, said: “This careful clinical investigation suggests that we could improve the outcome for the 500 people who go to hospital with a heart attack each day in the UK.

“By using a drug already commonly given to transplant patients, their findings can be immediately tested in heart attack patients.  We look forward to the outcome of Professor Spyridopoulos’ trial.”

This research was funded by the British Heart Foundation and the National Institute of Health Research (NIHR) Newcastle Biomedical Research Centre.Press release courtesy of The British Heart Foundation.

Reference: Stephen Boag et al, T lymphocytes and fractalkaline contribute to myocardial ischemia/reperfusion injury in patients. Published online in The Journal of Clinical Investigation July 2015. DOI: 10.1172/JCI80055

In ‘Self/less,’ a man cheats death by transferring his mind to another body — here’s how close we are to making that reality

July 16, 2015

sergio canavero

In the new movie “Self/less,” which comes out Friday, July 10, a wealthy man dying of cancer (played by Ben Kingsley) cheats death by transferring his consciousness to the body of a younger man (Ryan Reynolds).

Thanks to the help of a secretive doctor and a lot of money, all the procedure requires is going for a short spin in a device that looks like an MRI machine.

Of course, the mind that originally belonged to the younger body still exists — which creates some problems for the mind that has now taken up residence there — but the new body’s original mind can be “suppressed” by taking special pills.

At first glance, this sounds like a pretty sweet deal. But just how far off is this kind of technology? Do we know enough about the brain to even begin to undertake such a procedure?

In recent years, scientists have been making impressive strides toward understanding and manipulating the brain: We have rudimentary technologies for listening in on, and even altering, the mess of complex activity in the three-pound hunk of flesh in our skulls. Scientists have even developed methods for probing the brain using light, a technique that has been used to implant or erase memories in mice.

But as far as transferring the brain’s consciousness, a concept scientists still have yet to completely understand or define, we’ve got a pretty long way to go.

Problem #1: Everybody has a different brain

While the movie makes it seem like we could simply swap memories between two people, it’s not that simple: We’d have to also transfer the process our brains use to generate thoughts, Wolfgang Fink, a neuroscientist and roboticist at Caltech and the University of Arizona, told Business Insider.

“The reason why [the “Self/less”-style body-swapping procedure] isn’t really possible is that everybody has a different brain,” says Fink. “You would have to transfer not just the memories, but the same thought-generating process.”

In other words, it comes down to not just what we think, but how we think. Each of us has unique mental hardware, which is why it’s likely not possible to simply download your consciousness onto another person, or to a computer, for that matter.

Scientists have developed computer systems modeled on this hardware called artificial neural networks, some of which can flip back and forth between two different “mental” states. “That is sort of the closest to what we saw in the movie,” Fink said — which is why the main character had to take pills to suppress the other personality coming through. You have these two competing personalities, and we can mimic this competition in software. But a computer may not be able to reproduce you.

Which brings us to our next problem.

Problem #2: We can’t just implant memories

For much of its history, neuroscience has been confined to passively studying the brain. But in recent years, a technique has been developed that allows scientists to actively manipulate its activity using light.

Known as optogenetics, the technique involves injecting a harmless virus (containing DNA found in glowing algae) into neurons in the brain, which causes them to produce a protein that makes the cells active in response to light.

By shining a laser onto these cells, scientists can essentially turn them on or off.

In 2013, scientists at MIT used this method to implant a false memory in the brains of mice. In the study, the researchers placed the animals in a chamber where they received mild foot shocks, creating a fearful memory stored in a brain region called the hippocampus. Then, by shining light on the neurons that encoded the shock memory when the mice were in a different environment, scientists made the mice “remember” getting shocked in the new place even though it hadn’t happened.

lab mouse mice ratFlickr/Global Pandora

The same researchers took things a step further in a study this past June, when they activated happy memories in mice that were behaving as if they were depressed. They lost their usual appetite for sugar water, and didn’t put up a struggle when picked up by their tails, for example. But when the experimenters shone light on the mice’s neurons that activated a memory of a happier time (which they’d found by peering into their brains while allowing them to enjoy some time with female mice), it “cured” the animals’ depression, the researchers said.

Of course, the studies were in mice, not humans. And implanting or altering a simple memory is a long way from transferring the entire set of thoughts and memories from one brain to another.

That brings us to our next problem.

Problem #3: There’s an alternative, but it’s even trickier

Rather than transferring someone’s mind to a different brain, it’s more likely we would transplant your entire head to a new body, said Fink.

An Italian neurosurgeon hopes to do just that. Sergio Canavero of Italy’s Turin Advanced Neuromodulation Group announced his plans earlier this year to perform the world’s first head transplant as early as 2017. He has already secured a volunteer for the procedure, a Russian man with spinal muscular atrophy, a disease that causes the muscles to waste away and is ultimately fatal.

TheWhyA clip from the Konami video game “Metal Gear Solid 5: The Phantom Pain,” shows a man who shares Canavero’s exact likeness.

Canavero’s proposed procedure, which he calls the HEad Anatomosis VENture, or “Heaven,” involves finding a brain-dead donor whose body is intact, severing the heads of both the donor and the patient, and attaching the patient’s head and spine to the donor’s body.

While Fink thinks the head transplant may actually happen on schedule, critics say there are a lot of challenges to overcome.

For one thing, the new body’s immune system could reject the head, just as your body can reject a transplanted organ.

In addition, once the patient’s spinal cord is severed and attached to the donor body, the body may end up paralyzed. Canavero has claimed he can get around this problem by using a very sharp knife to make a clean cut that would allow remaining nerve fibers to repair the incision, but this has yet to be demonstrated. (In the past, scientists have performed head transplants with monkeys, but the animals only lived for a few days.)

The takeaway

Even if the procedure were technically possible, it brings up a host of ethical and philosophical issues. Should you be able to inhabit another person’s body? Would you still be you? Furthermore, any life-extending technology would undoubtedly be very expensive, so would it be fair that only the rich could have access to it?