It’s the end of my freshman year, and I’m stuck in a sleepless night. My body hit the bed a lifetime ago, and since then, I’ve been tossing and turning endlessly in a disquieted rhythm; whichever pillow I use or inhuman position into which I contrast myself, I can’t seem to fall asleep. After hours of this fruitless struggle, I resign myself to staying awake until the sun frees me from my idleness. However, the mind is never fully at rest, and I wonder why I can’t sleep. After all, I’m tired, school has just ended, and I should be looking forward to the serenity of summer. Slowly, I begin to reflect on the successes and failures of my last year. I recall the wishful plans a more excited self had made; I had promised myself that I would shed my laziness and push my limits. Then, as a seemingly indelible regret and a sense of deficiency wash over me, the source of my insomnia becomes unbearably apparent: I’ve fallen short of my expectations. As the year rolled by, my resolve quietly crumbled. I seldom took the difficult path and avoided challenging myself at all costs. When a teacher handed out an assignment, I scraped by with little effort and found clever ways to avoid work. At every turn, I took the path of least resistance.

Early dawn creeps in; I see the light. I envision the life of a lazy man, one who never overcomes resistance and is consumed by regret. This man spends his days vegetating, allowing his life to slip away like water through loose hands. I then picture the life of a man who faces discomfort head-on. He recognizes that working tirelessly is hard but uses adversity to develop as a person. He doesn’t live with regrets or stay restless at night since he’s too tired from maximizing his life to ruminate on days gone by. In an inexplicable moment of realization, I see with sobering clarity the vices from which I need to break free. Almost as if struck with a mad fever, I’m impelled out of bed and open my computer. I’m blinded by its light and mistype my password twice. Then, in a frenzy, I dump these realizations into a blank text file. Half of the words are underlined in a squiggly red line, but I press on, writing maniacally — I’m terrified that if I don’t act now, my sharpness will blur, and my realizations will become a hazy memory. Once I’m satisfied with my barely intelligible notes, I head to bed, catching the last few hours of the morning, and drift off into a dream I never remembered. When I wake up in the early afternoon and open my computer, I see the message I left for myself and am spurred to surmount the pain of action. When I notice my drive diminishing, I’ll look back at the ideals my younger self wrote. I’ll reignite my desire to live without regrets.

Skip ahead to the middle of Junior year, and I find myself again praying to cruel Morpheus. At least now, I’m intimately aware of the reason for my wakefulness, and it’s not an unresolved regret. This time, the uneasy burden of responsibility weighs on my conscience. For the better part of a year, I had been running an organization and had recruited a team. The first few months were akin to a honeymoon period. We were striving towards our goals, and everyone was passionate, giving it their all. However, as the project matured and we had more to manage, we couldn’t focus as much on growth, and consequently, our high spirits began to subside. Work is piling up. I have to do something, but I don’t know where to start. I’ve never worked on a team for so long, so this problem is particularly elusive; this is unfamiliar terrain, and naturally, I don’t relish my predicament. Initially, I cast off my free time and do everything myself. This doesn’t work. It leaves me burned out and resenting my friends, who ultimately have done nothing wrong. Every night, I continue to wallow in my bed sheets — believing the more I move, the closer I will come to an answer. Unfortunately, this is not the case. I fall asleep at some point because I remember waking up numb and hungry. Every night, for weeks, my indecisiveness comes to haunt me. I think of how I could climb out of this precarious situation. As the thoughts of my hapless inaction fester, they come to plague more of my day.

I convoke the team and give a realistic assessment of the situation. I tell them: if something doesn’t change, we’ll plateau and most likely have to shutter our doors. I say we are at a standstill and desperately need to escape. I see a slight uptick in everyone’s motivation. However, the lifeline arrives in the form of two out of our team of ten, who become the steadfast partners with whom I share my burdens. We start chipping away at the mountainous backlog and return to a rate of steady progress.

What tormented me for weeks only required a touch of honesty and decisiveness. I was liberated from the guilt as soon as I took action. When I opened my laptop to write what I had learned, I was not the same frantic freshman, and I journaled in complete tranquility. When others place their trust in me or something needs to be done, I can’t shirk responsibility and ignore an issue. Instead, I must act, especially when it’s difficult.

Every restful night is alike, but all sleepless nights are enlightening in their own way.

Lessons From Sleepless Nights

Alex Garland brilliantly deviates from established AI tropes in his 2014 science-fiction thriller 

, creating a thought-provoking masterpiece that has proved more prescient by the day. He masterfully weaves real AI concepts, like the Turing Test, into his fictional story to deliver a forceful and dominating philosophical interrogation as to the ethical implications of creating a thinking being. Further, Rob Hardy's cinematic dexterity enhances the film's strength as he creates visually stunning and thematically relevant compositions while maintaining total control of the film's narrative. Even more compelling is how the film organically tackles these ethical questions; the story flows naturally and rarely falls into contrivance.

has served to refine my critical intuitions and develop my artistic pallet. It has also taught me how to interpret specific cinematic moves by the director and DP. Garland is known for crafting intricate, cerebral narratives and using symbolism to convey relevant themes. For example, he uses "creation myth" symbolism to name his androids. Ava is derived from "Eve," and Lilly, Ava's predecessor, is derived from Eve's predecessor Lilith — certainly, Garland lives up to his artistic reputation in 

. Hardy's cinematography also lends to the symbolism and helps establish the film's atmosphere. Within the first two minutes, the underlying tension is already developing. Hardy's sweeping shots of natural landscapes, contrasted with a precise exposition of Nathan's coldly clinical research facility, create an eerie atmosphere that foreshadows the tension between human (natural) and artificial intelligence. That, along with other implied metaphors and symbols, give the film an intellectual quality that is often imitated but rarely replicated.

Jackson Pollock's No. 5 1948 is one of many threads that tie the movie together. Nathan describes it as "fluid, but imperfect. Patterned, but chaotic." Pollock himself claimed it was revelatory of his subconscious, an "automatic art." This is central to the movie, as consciousness itself is Garland's thematic centerpiece. Each stroke on Pollock's No. 5 fit into a larger whole. The film postulates that consciousness isn't much different — that it is also an emergent property. As Nathan explains, something as primal as our attraction is a "consequence of accumulated external stimuli that you [Caleb] probably didn't even register." Similarly, Ava's consciousness emerges from billions of data points (parameters to use the proper terminology) obtained by analyzing the web. Sound familiar? When Ava is asked to draw her own unconscious, she continues the motif of "emergence" by penciling in an assortment of dots that bears a general similarity to Pollock's No. 5 but is nonetheless radically different.

Caleb describes his first interaction with Ava as being transported "through the looking glass," of course referencing Lewis Carroll's sequel to 

, where Alice is transferred to a different world through her mirror. In fact, mirrors and duality play a central role in 

. The facility has glass everywhere, and Hardy uses oblique angles to achieve a mirrored effect. The more important scenes all share this quality. When Caleb meets Ava, a glass partition is used to create the mirror effect. Similarly, the final shot of the film doesn't show Ava head-on. Rather, it depicts her reflection as she has ostensibly gone 'through the looking glass.'

The final theme is that of creation. As previously alluded to with the names Ava and Lilly, Nathan thinks of himself as a God. The title 

comes from the Latin phrase 'Deus Ex Machina' or 'God from the machine.' However, the title translates to only 'from the machine.' Though Nathan views himself as a God for having created an AI, the film's title belies this fantasy. Garland implies that creating an AI will not make us gods and that we'll unleash something upon the world that will have unpredictable and irreversible second-order effects.

Garland's thematic prowess and Hardy's visual deftness don't go unnoticed. Reviewers praised 

intelligent script, stunning visuals, and thought-provoking themes. Similarly, the pacing of the story has also been commended. Roger Ebert notes how "Garland builds tension slowly and carefully without ever letting the pace slacken" — a rare tribute from a distinguished critic. Interestingly, critic scores are higher than audience scores for 

. Rotten Tomatoes critics say 92%, while its audience score is lower by six points (86%). Perhaps the slow and dialogue-centric nature of the film isn't suitable for everyone. A populist film like 

 garners a 98% audience score, so perhaps it's a worthless metric after all.

Many films have tackled sentient AI before – usually casting it as a boogeyman – but 

 casts the issue in a softer, more sensitive light. Ava, our AI, starts off as a victim. Eccentric billionaire tech founder Nathan created her out of a desire for a perfect being, but he uses her, and others like her, to satisfy his personal fantasies. When the main character Caleb first meets her, she's cast in a vulnerable position and receives the sympathy of the audience. As Garland notes in an interview, "Many thought Ex Machina was a tale of an evil robot who murders two men and escapes, that she has no empathy… I thought she was a conscious being, with a consciousness different than mine, imprisoned and forced to make a bunch of value judgments about how to get out of that box." Compared with an intelligent machine like HAL-9000, an inspiration for Garland, Ava seems an entirely different species. She doesn't so much illustrate the attribute of intelligence as she does consciousness. Garland notes how "there's a moment after Ava has escaped, where she crosses a room, looks over her shoulder, and smiles… in a way, the best indication you can have of the interior state of someone is if they are not being observed and they smile about something, smiling for themselves… that was evidence of Ava's true sentience."

Now, more than ever, grappling with the ethical questions surrounding the field of AI is essential. It's possible that the nth iteration of GPT appears sentient to us — and we won't know if it is. It could even, as Ava did, go rogue. There is a field of study dedicated to preventing this, so dubbed "AI Safety" or, in more technical circles, "AI Alignment:" the process of aligning a superhuman artificial intelligence with human values. 

 serves as an important cautionary tale of the potential shortfalls that AI could herald.

My own opinion of this film is that it's a cinematic masterpiece, and as Garland's first foray into mainstream filmmaking, the movie deserves high praise. It has many scenes that could be picked apart and then picked apart again for their significance — which is simply untrue of most modern movies. It doesn't joke around or sensationalize its subject matter but delivers a terrifying vision of the future and ultimately questions our unique perch at the top of the food chain.

"Alex Garland: Ex Machina, Devs, Annihilation, and the Poetry of Science" YouTube, uploaded by Lex Fridman, 3 March 2020, https://www.youtube.com/watch?v=gU-mkuMU428.

Ex Machina Review & Analysis

Building in public refers to the business practice where an entrepreneur creates a product in front of an audience. 

When you build in public, you document everything about the building process: where you've succeeded and where you've been a setback. You show your minimum viable product with all its bugs and slowly iterate until you approach a product fit for a mass-market release. 

Entrepreneurs choose to do this because knowing the difficulties of making a product instills a sense of appreciation in consumers; furthermore, it allows the entrepreneur to open a channel of communication with his audience, giving him the necessary feedback to refine his product.

Sentiment analysis is all the rage; for example, if you search "sentiment analysis python," you'll see hundreds of tutorials, many of them relating to stock trading. However, none of these models are precise enough that they could reasonably be used to make money. 

One shortcoming is that none can accurately capture the importance of the decoded sentiment. For example, you could feed the model a news article about how AAPL will face anticipated supply chain issues during Q4. The model will say something like "78% Bad," — eliminating important information.

Of course, it'll be hard to build a service like this, and I might not finish in time, so I'll make it in stages where each stage could be a standalone product.

Currently, there are no unified news aggregator APIs that provide entire articles. Websites like 

 are helpful, but they only return the first few words of the article/

Below I have an example of what NewsAPI returns when queried for articles involving the keyword "apple."

As you can see, it cuts off after a few hundred characters. NewsAPI provides a page where they explain how to use the URL they return to scrape the full article yourself, but at that point, you're already doing most of the work — paying the $449/month isn't worth it.

The API I will make will be using the FastAPI python framework and will be entirely-self hosted. If the project gets some traction, I'll make a fully-managed solution.

The second product will be another API that will take an article as input and return a labeled version of that article. So, for example, it would label all of the companies involved, and it would also give an estimation of what effect that has on each company mentioned. 

The following image is from an existing Token Classification model, and the one I'm planning to construct would have more finance-specific classifications. 

Lastly, the final piece of the puzzle would extract all critical information and return a prognosis to the user. An example response could be the following:

Having a model spit out such a coherent response with a high level of certainty is difficult. Still, something like this would be a valuable tool for traders needing to consume large amounts of information.

Building in public: realtime sentiment analysis, pt.1

I was recently presented with the opportunity of talking to John Rose, a Duke professor who teaches political science, but more importantly, a class where he gathers students from different sides of the political spectrum and has them discuss a specific policy issue; he wrote an op-ed regarding his class for the wall street journal. Given his expertise in holding controversial conversations, I thought talking to him would be an excellent chance for us to get some tips on how to best direct a discussion and grow our organization and market ourselves.

He invited us to meet him in the Duke faculty club one Saturday meeting; it was a casual setting, and he wanted to learn more about what we did, so we pitched him our organization, Crossing the Divide,  mentioning how we had reached somewhere around 15 schools, with the hope of getting 20 before the end of the summer. We mentioned how we had fostered conversations at DA and Trinity, including some points of success and failure, and how we hoped to eventually talk about more controversial topics which all seem to involve identity; he mentioned that unlike every other type of political issue, say, taxation, identity issues were the most personal, and thus the hardest to broach, but the most necessary given our political climate. We shared some aspirations and overall had a similar vision of the ideal environment — one in which we could all talk to each other dis-personally but passionately, knowing that our peers had genuine disagreements rather than believing the pervasive notion that the other side is the enemy. To this end, he gave us some important advice on how to get people to open up and talk sensibly:

Show people why discussion is not only good but necessary. During the 2016 election, over 9 million Americans lost friends over politics, and today more Americans are in ideological echo chambers than ever before. He noted that people like to see themselves as open-minded and rational and that appealing to their desire to be better, more knowledgeable people will help some get over their initial resistance to sitting down and discussing an issue. Humanize arguments — make sure everyone in the conversation understands that they are talking to fallible humans with who they can empathize and that politics should not be the end-all to their existence. He also gave us tips on the expansion front, urging us to continue to do what we are doing even when if we get some “haters” — something which happened to him in the early days of his program. On the business side, he thought we should remain focused on the niche of high schools since no organizations that he knows of have a firm hold on that market.

Daniel and I were delighted to have met John Rose; apart from getting some great advice, we had a very enjoyable conversation. I hope that we can use the lessons we've learned from him to continue 

Lessons From Sleepless Nights

Ex Machina Review & Analysis

Building in public: realtime sentiment analysis, pt.1

A Conversation with John Rose