Innovation's Speed Asymmetry/ The Quest for Synthetic Life/ RISC-V to Turn The Chip Industry on Its Head?/ Using AI to Unearth New Smells/

Innovation's Speed Asymmetry. This week’s collection of articles is the perfect summary of the potential and the risks associated with the incredible advancements coming from what I call the deep tech approach to innovation.?

Thanks to two convergences happening at the same time, we are dramatically expanding the option space for innovation. On one side the convergence of science, design (intended as problem-solving) and engineering allows for addressing problems that seemed intractable. On the other side the convergence of technologies around matter and energy, cognition, and computation, as well as sensing and actuation make the solutions to these problems possible, thus expanding the option space by orders of magnitude and enabling 10x innovation vs. 10% innovation.?

Additionally, through the utilization of the Design-Build-Test-Learn cycle, the speed at which these innovations are being made possible is similarly reduced by order(s) of magnitude. Generative AI allows for new designs, the prototyping capabilities allow for the generation of data through testing, leading to improved learning, and forming a virtuous circle that can lead to breakthrough solutions in a much faster fashion.

But this approach to innovation does not come without questions or risks, as broadening the options space, implicitly means entering unchartered territories, for which we don’t have (easily) the ethical and the safety answers to be able to operate in the enlarged option space safely…?

Going through today’s edition of the Antidisciplinarian vividly reminded me about both the potential of the deep tech approach to innovation, but also about its risks.

We have the manipulation of matter by making DNA and transforming it into “synthetic life”, which requires multiple disciplines to happen (and clear safety and ethical boundaries with it), we have the design of smell leveraging computation, and we have new computational and coding paradigms, we have ethical questions related to ChatGPT, how to translate text into music and an attempt at using non-used neuronal bandwidth to create non-invasive brain implants…

The pace at which all of this is happening is not to decrease, quite the opposite… and we need instruments to be able to deal with it adequately as humanity. The way we are going to address this asymmetry between the (ever-increasing) pace of innovation and the way and speed at which we as humanity will respond to that will massively impact our collective future. There is no easy answer, and there are people that are way more qualified than I am to discuss this (as you will see in the next weeks).

As far as I am concerned, three things are going to be key to successfully addressing the asymmetry in speed:

1)????The role of individuals as moral actors will become more important, as it will be more and more difficult to rely on society (given its reaction time) to elaborate on technology-related ethical questions. I think important for individuals to develop a strong individual moral compass and be thought about how to do so.

2)????The role of artists in pushing the boundaries of what is possible and anticipating the related ethical questions will become even more precious than it has been over the past centuries.

3)????Given the expansion of the option space and the concomitant collapse of the time of innovation cycles, it will be mandatory for all innovators to develop a dialogue with the public, so to create a joint answer. The worst thing that could happen to us would be a separation of emotional (hence irrational) and scientific (hence rational) narratives around what is going on.

On a different, yet related, topic:??Niko McCarthy essay below ?on DNA synthesis is one of the best pieces I have read on the topic for a long while, and I agree 100% with all his observations. Ultimately, if we want to move from an exploitative to a generative paradigm, we do have to go through the DNA to translate its potential into reality, which is exactly what we are trying to do at?Officinae Bio . And while the potential is huge, there is still plenty of work to be done to make it happen, in a safe manner.

Making DNA: The Quest for Synthetic Life

The?Black List Printer ?- an art installation by?Georg Tremmel and Shiho Fukuhara ?that creates only “forbidden DNA sequences.”

A deep dive into the past, present, and future of DNA synthesis and how we “stand at the precipice of newfound capabilities… limited only by imagination.” The last decade’s advances have made it easier “by at least an order of magnitude” to build custom DNA. Could “synthetic organisms” one day help us?terraform Mars ? “New-to-Earth lifeforms could make almost anything.” But “biosecurity experts ?worry about the risks” of DNA synthesis tech becoming widely accessible.

The authors argue that “democratizing” DNA synthesis is necessary to enable “fresh minds, with bold ideas [to] enter the fray” and argue for “a grand scientific mission, with balanced risks and uncertain outcomes.”

News items:

Moderna, Becoming a Buyer, Acquires a Synthetic Biology Specialist

Flush with cash ?from its Covid-19 vaccine “windfall,” Moderna has acquired?OriCiro Genomics ?for $85M. Moderna CEO Stéphane Bancel?says ?the acquisition gives the company “best in class tools” to manufacture “a?key building block ?in mRNA.”

These Simple Design Rules Could Turn the Chip Industry on Its Head

RISC-V ?- an open-source chip building?instruction set ?- “could radically change the chip industry by making “design more accessible to smaller companies and budding entrepreneurs.”?Intel has pledged $1B ?to “develop the RISC-V ecosystem,” and “there are already billions of RISC-V-based cores out there, in everything from earbuds to cloud servers,” says?Mark Himelstein , CTO at?RISC-V International .

RISC-V may eventually break Arm and Intel’s stranglehold on?the chip industry , but “it’s not yet clear whether RISC-V designs will supersede them.” Additionally, not everyone is a fan of RISC-V’s “open philosophy.” Amid?rising geopolitical tensions , lawmakers accused RISC-V of potentially causing the US to “lose its edge” in semiconductor design to China’s benefit. The RISC-V International non-profit?relocated to Switzerland ?in 2020.

News items:

Roboticists Want to Give You a Third Arm

Thanks to “unused bandwidth” in your neurons, you may soon be able to augment your abilities with an “extra robotic limb.” Invasive brain-machine implants aren’t a viable option for most individuals, but?recent advances ?in electromyography ?(EMG) may provide “a noninvasive way to pick up brain commands from outside the skull.”?

This Startup Is Using AI to Unearth New Smells

The $30B global fragrance industry increasingly faces both?climate-related ?and?ethical challenges ?to how it sources some of its most memorable scents. Sandalwood trees are?endangered ,?saffron ?and?vetiver ?are vulnerable to “geopolitical turmoil,” and natural musk requires?endangered animals ?to be?captured or killed .

Google Research-affiliated ?startup?Osmo ?- recent ?recipient of $60M ?in initial funding - is “trying to teach computers to smell” and use AI to “create the next generation of aroma molecules for perfumes, shampoos, lotions, candles, and other everyday products.” Osmo co-founder and founder/MD at?Lux Capital ,?Josh Wolfe , says “We see Osmo as a rational design business model where people want a very specific odor and we design [and license] the chemicals.”

News items:

Apple Needs to Take Its Time With Augmented Reality

Apple’s “long-rumored AR headset” may be announced as soon as this spring and could launch in the second half of 2023. But maybe Apple shouldn’t be in a rush. The “biggest names in the industry” have all?swung and missed ?at AR/VR hardware. If the “world’s most well-capitalized design studio” fails at AR, “that reality would?really?bite.”

ChatGPT’s Mind-Boggling, Possibly Dystopian Impact on the Media World

The furor over ChatGPT shows no sign of slowing down. The generative AI model inspired a?widely ?covered ?takedown by “underground rock icon”?Nick Cave ?- who called ChatGPT “a grotesque mockery of what it is to be human” - as well as a?Guardian editorial ?by a self-proclaimed “copywriter” who is “pretty sure AI is going to take (his) job.”

Meanwhile, established news outlets like Buzzfeed are embracing the technology. CEO and founder?Jonah Peretti said, “We see the breakthroughs in AI opening up a new era of creativity that will allow?humans?to harness creativity in new ways with endless opportunities and applications for good.”

But has the backlash already begun? CNET?was revealed ?to be using an “AI-content farm” to publish stories that turned out to be?“full of errors” ?and?“teeming with robot plagiarism.” ?Even influential newsletters?aren’t immune . After having “passages lifted directly” from his?Big Technology Substack , journalist?Alex Kantrowitz ?- who remains “super bullish on generative-AI” - said models like ChatGPT’s “darker side will be difficult to tame.”

(Wondering if an AI wrote?this?newsletter? This?free browser plugin ?claims to identify “AI-derived copy with 99.2% accuracy.”)

News items:

Google’s New AI Turns Text Into Music

Need an AI-generated original (?) soundtrack to go with your?imaginary film stills ? Google’s?MusicLM ?was trained on “280K hours of music” and “can generate minutes-long musical pieces from text prompts.” You can?hear samples , but you’ll have to wait to turn your words into music - citing?copyright concerns , Google isn’t making MusicLM public (yet).

penAI Has Hired an Army of Contractors to Make Basic Coding Obsolete

OpenAI is hiring contractors worldwide to help teach its AI models to better “translate natural language into code.” OpenAI’s existing?Codex ?product was trained on?code scraped from GitHub ?and is currently used to power the Microsoft-owned software repository’s?Copilot ?tool - a sort of autocomplete for coders.

Semafor’s?Reed Albergotti ?believes that teaching AI models how to write basic code could be “as transformative to [software development] as heavy equipment was to the construction industry.” Similar to generative AI models like DALL-E 2, “creative people with… no coding experience” could be able to build “everything from websites to video games” by describing their ideas in natural language.

Andrej Karpathy , former director of AI at Tesla,?tweeted , “The hottest new programming language is English.”

News items:

I Edited Human DNA at Home With a DIY CRISPR Kit

Famed ?biohacker ?Jo Zayner ?wants to “democratize the tech” behind gene editing. Their company,?The Odin , offers “DIY genetic engineering kits” and online courses on how to use CRISPR to edit human DNA in the comfort of your own home.