Femtosecond

On how thinking about the future and its technologies should maybe lead us back to the present more often

Thanks for reading my newsletter! The breakdown: first a personal essay, some of my recent writing, then some thoughts on the media that I am consuming and finally some vegan food and recipe recommendations. All opinions expressed here are strictly my own. 

If you are here because you like my writing about science or my Instagrams about cooking, you may not be interested in every essay in this space, but please do stick around until I loop back to whatever it is that we have in common.

Find me on Twitter and Instagram. I’d also love it if you shared this letter with a friend.

FEMTOSECOND*

“Do you think that, like, the average chemist does think that AI and machine learning will be a part of chemical discovery in the next 10-20 years?” I asked a chemist who squeezed in talking to me between two presentations at a busy conference in a sprawling San Francisco convention center. 

Huddled over my work laptop in the back of a press room that simultaneously felt too empty and too busy, we had been talking about using artificial intelligence to put every chemical compound that could ever exist on a digital, multi-dimensional map.

“Yes. And I think that's changed in the last six months,” he said somewhat tiredly but with more specificity and certainty than I had expected. 

We kept talking about how so much of how researchers do chemistry and what everyone thinks being a chemist entails is bound to change almost as if it was a given that we will wake up in that brand new world of artificial intelligence boosted science tomorrow. Of course, being a scientist of any sort, and especially one that speaks with the media, often means excelling at caveats so we spoke about many if’s and maybe’s as well. But I was still struck by what seemed to be a sincere belief, one that I did also witness in other chemists as the week went on, that we may be on the precipice of a big change. 

Though I have reported from conferences before, I was mildly overwhelmed by the dozens of thousands of chemists who were also attending the Fall Meeting of the American Chemical Society to present their research or take meetings with future collaborators and mentors. I scrolled through the buggy app that contained the schedule for the meeting so frantically that one of my thumbs got sore and every talk I recorded revealed a steady background click-clack of my anxious typing when I listened to it after returning to New York. When my editor called to check-in on day two of the Meeting I babbled about the chemistry of silica dust on the Moon for at least three minutes before managing to take a breath and make an actual point. 

“There’s a lot of chemistry here that I had no idea even existed,” I admitted a little too incredulously for someone whose job it is to keep track of new scientific endeavors. 

The whole five-day conference was organized around the tagline  “Harnessing the Power of Data”. So, much of that new chemistry that I was learning led me to automation and artificial intelligence, to robots and computers playing supporting roles in labs of all sizes, alongside chemists of all sorts. 

A big commercial company sent a researcher to present on a robot that tests new adhesives by sticking and unsticking them hundreds of times during the 8-hour workday. A pharmaceutical start-up boasted getting a cancer drug from “the bench to the clinic” in under a year thanks to a streamlined testing and discovery process that included multiple proprietary machine learning and AI programs. A computational chemist spoke of an AI “recommender” for making the best materials that can store solar energy or capture carbon dioxide. Another argued that methods for compressing data similar to turning a music file into an .mp3 can help chemists simulate the behavior of molecules tens or dozens of times larger than they ever could before. 

When I searched “machine learning” in the conference app it returned 274 pages of sessions with titles like “Helping Chemists Manage their Data”, “Data-driven Design of Energy Materials” and “Simulation and Data Science Approaches to Design Biologically Relevant Polymers and their Applications.”

“One of the things that we have to get out to kids is that chemistry has changed and is continuing to change. So if they got an image of what being a chemist is, it's probably wrong,” that one chemist told me when I asked whether future scientists should already be learning how to use machines as research assistants. 

The last time I had seriously engaged with chemistry was in high school so I could see his point in action throughout my time spent shuffling from talk to talk armed with a flimsy coffee cup and fighting the three hours of tiredness gifted to me by Pacific Daylight Time.In fact, a lot of what I saw at the Meeting would have struck the high school me as almost unimaginably futuristic, like the casual use of the gene editing tool CRISPR in so many presentations about food or medical chemistry or the fact that idea of robotic arms running hundreds of experiments from start to finish only prompted scientist to debate the best software to control them with. 

I kept thinking about the Jetsons and Profesor Baltazar and every single science fiction property where a computer talks back. Yet, actually doing my job and choosing which pieces of science to pitch to my editors as well as accepting their disinterest in commissioning stories about most of them, seriously grounded my excitement about the future of chemistry either being now or at least coming up tomorrow. 

This sort of tension between “I still can’t believe that we can do that” and “this isn’t new enough for the news section” permeates a lot of what I do as a science journalist. The psychological time necessary for letting go of shock about something like gene editing being real does not at all match up with the extra short timeline on which headlines’ grabbiness expires. And when I am not reading papers or attending a conference so the shiny, shocking new thing is hidden from my sight, I do routinely forget how futuristic my world already is. 

For instance, because I was a public school teacher at a time, I received the first dose of the COVID-19 vaccine fairly early on in the pandemic and have since gone on to get multiple doses and boosters. This vaccine was for all intents and purposes a remarkable breakthrough in biotechnology and though I did appreciate it as a promise of increased safety at an uncertain time, the implications for how quickly medicine is progressing mostly escaped me. Now, the mRNA technology that these vaccines brought forward stands a chance of being used to treat everything from bacterial infections to autoimmune conditions, rare genetic disorders and cancer but the wow factor of how they actually do that - by instructing our bodies to make medicines that we need all by themselves - never even caught on. 

CRISPR, a gene-editing tool that was recognized by the Nobel Prize in 2020, less than ten years after its discovery, is similarly something that shows up in the pages of New Scientist and similar publications with such regularity that we have style guidelines for how we explain it, but I have repeatedly encountered friends who despite being very educated and very online had never heard of it. This always surprises me, but maybe it really should not given the economic and ideological climate we live in.

Part of the mRNA vaccine’s downfall was certainly political, and with CRISPR part of the problem is in the economics of how it may reach the average person. The first genetic therapy based on CRISPR is set to be approved in 2023, based on years of trials within which more than 30 people were cured of sickle cell disease despite having previously been diagnosed with severe cases. This condition is named after the affected person’s misshapen red blood cells which can cause a lifetime of agonizing pain, anemia and problems with growth and vision. 

Previously, it could only be treated through bone marrow transplants so having a new approach that is more reliable and mostly includes blood transfusions could be a game-changer. Except that analysts expect that this therapy, made by Vertex Pharmaceuticals in the United States, will cost at least a million dollars per person, and the company has not announced any fixed price yet. In the US, thousands of people dealing with sickle cell disease, which disproportionately affects African Americans, receive health insurance through public programs. So, it seems unlikely that many of their lives will be changed by what should by all accounts be an unequivocally life-changing technology, and this economic barrier may even perpetuate existing racial and class inequities.

In medicine and medical science it is actually sort of a given that most patients, including the most wealthy, do not receive the care that is most aligned with recent breakthroughs in research. I was shocked to read one of my colleagues’ reporting about how the gap between what researchers know and what practitioners do is not uncommonly as large as 17 years. For context, some notable things that happened 17 years ago were the founding of Twitter, a fast declining platform now known as X, and Lordi, the trollish Finnish outfit, winning the Eurovision Song Contest. 

“There’s this huge chasm between how we develop our discoveries and how we actually get them into the world. People in universities developing things seldom include the end users. So they show that it works in very tightly controlled environments, but then it all falls down when they go out into the real world,” implementation scientist Rinad Beidas explained to New Scientist’s Corryn Wetzel. She also cited the lack of incentives for those people in universities to see their research make it past a publication in a high-impact journal as part of the problem. I have certainly encountered this in my reporting which is studded with could’s and might’s every time I write about a technology that should in theory someday be part of my life, like a plastic wrap that changes color when put on food that has gone unnoticeably bad or a soft medical robot that will deliver medicines into just the right parts of your body after you swallow it. 

I am not so jaded as to think that no breakthrough ever manages to leave the lab, but being a reporter certainly taught me that even a well-funded research project that ends up on the cover of Science or Nature may not have the financial support or people-power to become not just an invention but a product. Having this perspective does make events like the ACS Meeting feel somehow inherently fraught - the research on display amazes me and my inner teenager who loved science fiction, but the understanding of structure of science and economics of related industries quickly dampens that amazement. 

At the same time, thinking about how that teenage version of me would receive all this science did force me to confront that a lot of once futuristic technology did reach me and is now a part of my life. And I do benefit from features of that technology that may have seemed the most Star Trek adjacent at that time, like calling my partner through my watch if I lock myself out of our apartment or having the option to work remotely a few times a week. 

Even the idea of robots and automation is not just a dystopian vision of the future, but something that has happened already and is still happening. Researchers estimate that in the time period between 1990 and 2007 approximately 670,000 human jobs in US labor markets disappeared due to automation. The rise of large language models like OpenAI’s ChatGPT has for a good part of this year had many of my colleagues in journalism scared that a similar fate also awaits our profession soon.

Despite such fears which can be very intensely felt on the daily, thinking about our fate and our future seems to be something that human brains do not excel at. Functional MRI studies, which map the electrical activity in the brain, suggest that for most people their brains react to images of and ideas about their future selves as if that person is a stranger. Learning this made me wonder whether this is why it can be so hard to accept that, as the saying goes, the future is now -  the future self is an unrelatable stranger and the self we see ourselves as being now is someone we are often truly obsessed with. 

In fact, in a 2017 Institute for the Future survey, 36 percent of Americans said that they rarely or never think about something they may be doing 10 years in the future. “Compared with how often we think about our close friends and family—a near daily occurrence—we hardly give our own future selves a thought,” wrote the Institute's research director Jane McGonigal in Slate, underscoring the idea that the distant future is, for many, also emotionally distant.

In recent years, I have often thought that not being able to imagine a future radically different from a simple extension of our present is in part what drives much of animosity and fear among people who are all being exploited by our current economic system. Because we do not always have the capacity to imagine something different, we double down on the logic of scarcity and turn against each other instead of building solidarity across our differences. 

This lack of imagination also plays into how we envision the next step for futuristic technologies that are already on our doorstep. The same AI technology that may revolutionize chemistry has already been marred by ethical issues concerning copyright, racism and, as always, greed and a lack of transparency. 

Teachers are scared of it because the punitive, rigid and often overly demanding way we have set up schooling only leaves room for imagining how students will use the technology to cheat. Tech pundits are equally skeptical of those who claim that AI will save the world and those who have made warnings about its presumed ability to bring about doomsday their brand, and since both are tools for hoarding clout and social and other capital the skepticism is probably justified in both cases. 

So how can a technology truly feel like it has brought a capital-F Future within our line of sight if we already think, and in a self-fulfilling loop act as if it will never be all that different from what we have already?

I kept thinking about this as I was sinking a good chunk of July reporting on a material called LK-99 that briefly held the promise of being revolutionary. After a pair of uncharacteristically imprecise papers showed up on the preprint service arXiv claiming that LK-99 conducts electricity with perfect efficiency, or superconducts, at room temperature and ambient pressure, the Internet went berserk. Despite warnings from academics and other experts that have periodically seen similar claims be disproven over the decades since superconductors first entered the scene, X users that may have once been preaching the gospel of cryptocurrency or AI turned their attention to hyping up LK-99 instead. 

I was personally reluctant to write about the papers after seeing them on the arXiv, but after an X post that read “Today might have seen the biggest physics discovery of my lifetime. I don't think people fully grasp the implications of an ambient temperature / pressure superconductor. Here's how it could totally change our lives.” went viral my editors insisted that a news story synthesizing and explaining the skepticism of superconductivity experts was necessary. In that first story, most researchers I called would not go on record and one even yelled at me on the phone about how the arXiv needs better moderation for a truly impressive amount of time before I managed to ask a single question. But this story got the ball rolling and following the saga of LK-99, now being quickly synthesized and tested in labs around the world, became my full time job for almost two weeks.

It was exciting to see the scientific community spring into action so quickly and offer so much thorough information on social media almost as if they were giving everyone a window into how peer review and the scientific method are meant to work. But I was also dispirited by the more fanatic and less factual narratives that often garnered more clicks and likes than what career practitioners of science had to say. 

Blue-check owners on X spun stories about the researchers involved in the creation of LK-99 that eventually started to read as biblical apocrypha and they, mostly unfoundedly, claimed that if LK-99 works it could revolutionize everything from making commercially viable nuclear fusion easy to solving climate change. Once evidence started to mount disproving every single claim of extraordinary properties in this new material and most researchers considered the case on it closed, Lauren Leffer at Scientific American insightfully wrote about how even had it worked it may have not actually brought about such sweeping change so quickly. So why did it cause such a loud digital ruckus that the New York Times covered it in a style that almost belongs to page 6 rather than the science section?

After a flurry of papers explained away most of LK-99’s presumably miraculous characteristics as more mundane qualities, most X users moved on, and I could see people stop clicking on my reporting in New Scientist. But a few held onto the frenzy and started to fashion themselves as the only true believers. And it was only after they started attacking fellow science journalists that it dawned on me that they didn’t really need LK-99 to change tech as we know it. What they did need was for the story about it that they spun, rife with old cliches, to be proven true. 

A story about a group of underdogs going up against evil academic elites, often flavored with shades of orientalism as those first two papers originated in Korea, was simply more attractive than the truth of a whole community of experts getting to the bottom of a mystery in a short time. Ironically, many superconductor experts warned that claims like those that were made about LK-99 pop up every few years and often go ignored, but here were rabid commenters arguing for a different story that they think is secretly always repeating as well.

I do not want to imply that elitism doesn’t run rampant in academia and academic publishing or that it doesn’t matter where you’re from or that sometimes people really are genius underdogs in the eyes of a pretty crushing system we have set up for professional scientists. But it is a past-minded desire to focus on a very individualistic tale of a few against the many when all facts support getting behind a more collective, consensus-driven, communal effort and narrative in this case. And I hope any casual readers or X users that were on the fence did give at least some chance to the latter - science is at its best a team sport, and we should in fact champion that as its future.

There’s a quote from the futurist and researcher Roy Amara that a few scientists mentioned to me at the ACS Meeting, stating that “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” It’s a statement that feels intuitively true, especially if you grew up with the Internet and saw it spark a myriad of moral panics that then quickly fizzled out yet the Internet still managed to change how we do everything from buying books to voting in elections, and often not for the better. And it is a statement that once again makes me wonder about the kind of stories we like to tell ourselves about breakthroughs and about the future. Maybe we all just like temporary drama and the feeling of something new, some new win, and cannot relate to ourselves in the future who may be affected in some less sexy way, just trying to make it through the thing of their day like many of us are currently trying to push through our own right now.

The chemists who do believe that AI and automation are inevitable told me that one of the biggest challenges in working with them will be finding a way to gather and leverage data about what actually happens in labs when there are no big breakthroughs going on. In many cases, the procedure for synthesizing a useful molecule that finds itself in the pages of a prestigious academic journal is the twentieth or fiftieth one that a research group may have tried. For AI to really “learn” chemistry so that it can assist future human chemists, it will have to learn about all of those imperfect, suboptimal or even erroneous procedures. 

Successes and breakthroughs alone do not convey enough about the state of science for a machine, no matter how futuristic, to find a way to nudge it along. Machine learning algorithms can learn lots from journal articles, but abandoned theses, laboratory notebooks and informal research reports and notes shared between research group members may be what push it closer to being not just occasionally useful, but truly indispensable.

And so, if we let go of this need for drama, for a bang, for a splash that is actually quite foreign to the slow and incremental trot of science, we may be able to see that the future is very much here already, partly in the things we have built thanks to our imagination, partly in the bits of our days that we have prematurely decided to relinquish to something like a forgotten history. Though they may feel like strangers, the people of the future are already hiding within us, and we should maybe sometimes let ourselves be them already, even when they are boring.

Best,

Karmela 

*A femtosecond is one-millionth of one-billionth of a second and represents a time-scale on which lots of physical processes that are essential for chemical and biological processes, like photosynthesis, happen. Many processes that happen in femtoseconds are blips too short for us to even imagine yet they are crucial for our continued, yet much slower, existence.

Do you like Ultracold? Help me grow this newsletter by recommending it to a friend, sharing this letter on social media or becoming a subscriber

ABOUT ME LATELY

WRITING

I am genuinely quite proud of my coverage of LK-99 for New Scientist even though this was some of the most stressful and most rushed reporting I have ever done. But no matter how quickly I had to work, I was still reminded of how satisfying it can be to put together pieces of a story, something I do not necessarily get to do when I am reporting on one study at a time, and also that I do have the ability to get in the zone and deliver under pressure when I really need to. That little boost of self-confidence in my reporting and writing skills meant more to me than any number of clicks on the website or the unusually long word allowance I got in print (two whole pages!). 

I also quite liked reporting this story about using bacteria-killing viruses as nature-made nanobots for food safety. There are companies that sell these already, yet another example of how the future is here.

READING

One of my biggest downfalls is that I am a writer with no reading discipline so I am still very slowly working through Monolithic Undertow: In Search of Sonic Oblivion by Harry Sword, a book about a very broad history of drone and drone adjacent music. I continue to be annoyed by the pretentiousness of the narration in this work, but I also cannot let myself quit it because, so far, its zig-zagging path from Indian ragas to deconstructed jazz and tape music, to pre-Velvet Underground John Cale has been absolutely fascinating. 

This essay by Emma Specter in Vogue about violence, body image and Barbie that was certainly written during a time when the movie of the same name was at peak frenzy but its contents remain relevant and emotionally resonant. She writes:

…the fact remains that Barbie's body was the first one I learned I could control at will without worrying about what anyone would have to say about it. Isn't that what dolls so often do, after all; give us a trial space to play at the messy and constant work of womanhood? Was Barbie's body a proxy for my own, and if so, what does it mean that after the age of ten or so, I only showed her my love via Sharpie-scrawled, bobbed-haired "improvement"?

And I did really cut at least one of my Barbie’s hair short when I was a preteen.

LISTENING

The self-titled record by Townes van Zandt because someone played Dead Flowers at a camping trip I went on back in May which led me down a van Zandt rabbit hole months later. And because I am a sucker for a brutal kind of sadness that he still manages to make sound like poetry. 

The shouty, new-wave-y post-punk of Ausmuteants’s Band of the Future and, on the opposite side of the noisy spectrum, the slow, monotonous electronic soaring, twinkling and scratching of Tim Hecker’s No Highs.

WATCHING

My mom visited Brooklyn for the last week of July and the first week of August so we made seeing Barbie together into a small occasion which was very fun even if the movie did slightly underwhelm me. I am not the first to point this out, but as much as the film is visually impeccable and reaches levels of camp that are quite uncommon in mainstream films these days, it starts to feel underwritten about a third of the way in. The hilarious dance numbers and setpieces are just fun enough to distract from the flatness of most of its non-Barbie and non-Ken characters as well as a lack of a coherent political point of view. Certainly, this is a movie that aspires to a feminist message and is not shy about wanting to be labeled as feminist (maybe because conservative backlash helped its marketing) but its punchlines reminded me of reading budding fourth-waves-ers on Tumblr in the 2010s. This may be a function of who I am as much as of who bankrolled this film. Mattel was never going to support a radical critique of gender or heterosexuality (notably, the film is very queer-coded without including any explicitly queer characters or storylines) or capitalism and how they intersect with what may have once only been called women’s issues. And the average viewer may be significantly less salty about that than me. But hey, I wore my pink outfit and my pink makeup and still made the trek to the movie theater during Barbie’s opening week, and that may speak to its power, and how much less its philosophy and its merits matter, more than anything I could write here.

For a team-building event at work, I watched Marcel the Shell with Shoes On and participated in a very loose discussion of the soft-spoken mollusc in the office later. Similar to Barbie, the film felt like it belonged to a past era of the internet, when we all knew a little less, had been trained on less content, and were not quite so jaded about manufactured cuteness. Marcel is partly a film about the Internet and the perils of virality and I could sort of appreciate what it had to say about that.  I also really loved Isabella Rossellini’s voice performance. But everytime Marcel tried to deliver a beautiful truth about the complexity of life or struggles with personal growth I just could not quite get past the performed earnestness of it all.

My partner and I have also been watching Steven Soderbergh’s miniseries Full Circle and I have mostly really been enjoying it. This is a grimey New York show that brings together organized crime, celebrity chefs, postal inspectors and Guyanese dark magic into a pretty suspenseful plot full of people so irritating that they are actually believable. It’s a mashup of Thomas Pynchon, Elmore Leonard and the way New York’s dodgier parks feel at night, all put together tightly and coherently by Soderbergh’s skillful and spars direction. I hope the ending, which we have not watched yet, will not disappoint me.

EATING

I cooked quite a bit while my mom was visiting, but we also really indulged in meals outside of the house throughout her birthday week, including getting arepas at Rockaway Beach and attending a tomato-themed party where the amount of tomato-centric dishes on display was nothing short of impressive. We also made a trip to Chinatown for dim sum and to the incredibly well-executed and creative Mexican cuisine at Oxomoco in Brooklyn.

The best thing I ate in San Francisco was a yuba salad with a miso dressing at a vegan sushi spot called Shizen that a friend of a friend graciously recommended before I flew out. 

As the summer has been extremely sunny, and the risk of a COVID-19 infection still remains a real concern, I attended quite a few picnics in the last few weeks. One notably featured my friend Nabil’s incredible homemade seitan and another giving me the excuse to make a grilled peach and marinated tofu panzanella with great Italian bread from a local bakery and gorgeous heirloom tomatoes from our neighborhood farmers market.

On the dessert front, I was proud of this cake that included a vanilla batter and chocolate and coconut caramel frostings, and I garnished it with a chocolate ganache and chocolate dipped potato chips. I made it for a friend’s birthday after he voiced a penchant for mixing sweet and salty, something that I personally believe in too. And I was immensely overwhelmed with joy when a double date with my best friend and their partner landed my spouse and me at the dessert tasting at Hags, a distinctly queer high dining experience full of friendly and familiar faces that always feels like such an un-guilty treat.