AI copilots for software development work very well

With Github Copilot, software engineers can achieve much better productivity. Engineering teams need to embrace this tool ASAP.

This will make it difficult for any IDE to compete with VSCode. One of many ways AI could cement Microsoft’s monopoly on B2B SaaS.

Highlights

This paper presents results from a controlled experiment with GitHub Copilot, an AI pair programmer. Recruited software developers were asked to implement an HTTP server in JavaScript as quickly as possible. The treatment group, with access to the AI pair programmer, completed the task 55.8% faster than the control group

— arxiv.org

Hyperstition: something that is true if everyone believes it

A useful term in a time characterised by meme stocks and bank runs.

Highlights

Hyperstition is a positive feedback circuit including culture as a component. It can be defined as the experimental (techno-)science of self-fulfilling prophecies. Superstitions are merely false beliefs, but hyperstitions – by their very existence as ideas – function causally to bring about their own reality.

— www.orphandriftarchive.com

The US government made $15B from the 2008 bailouts

Not a lot of money, but it’s important to understand when considering future banking crises. Keeping the system together is always going to be worth it, but it’s especially encouraging to note that the taxpayer was made whole.

Highlights

The US government made $15B from the 2008 bailouts.That's 0.6% return on investment which is kinda poor but overall the bailout was profitable for US taxpayers.

— news.ycombinator.com

A complete map of an insects brain

Parallel to our efforts to build artificial intelligence are our efforts to understand natural intelligence. This research is going to be increasingly important as we approach AGI and contemplate the nature of consciousness.

Highlights

Now, researchers have constructed a detailed map of the neurons and the connections between them in the brain of a larval fruit fly. With 3,016 neurons and 548,000 connections, called synapses, the result is by far the most complex map of a whole brain ever made.

Until now, scientists had only made this sort of map, called a connectome, for the brains of three organisms: two types of worm and a sea squirt larva. These animals’ brains only had a few hundred neurons each.

— www.smithsonianmag.com

Kevin Kelly on the future of technology

These are some of the most important ideas for understanding the current and future state of the world.

Highlights

I call our human made system of all technologies working together, the technium. Each technology can not stand alone. It takes a saw to make a hammer and it takes a hammer to make a saw. And it takes both tools to make a computer, and in today’s factory it takes a computer to make saws and hammers. This co-dependency creates an ecosystem of highly interdependent technologies that support each other. The higher the technologies the more intertwined, complex, and codependent they become. At this point in our evolution we need farmers to support indoor plumbing and plumbing to support banks, and banks to enable farmers, and round and round

23 other inventors created electric incandescent light bulbs prior to Thomas Edison. Edison is renowned primarily because he was the first to figure out the business model of electric lighting

while technology has gotten us into this mess (climate change) only technology can get us out of it. Only the technium (our technological system) is “big” enough to work at the global scale needed to fix this planetary sized problem. Individual personal virtue (bicycling, using recycling bins) is not enough. However the worry of some environmentalists is that technology can only contribute more to the problem and none to the solution. They believe that tech is incapable of being green because it is the source of relentless consumerism at the expense of diminishing nature, and that our technological civilization requires endless growth to keep the system going. I disagree.

— noahpinion.substack.com

How to save Rome

An interesting article with a few challenges for AI. I love the answer to the question for sending useful ideas back to Rome.

Highlights

What are four sentences that I could send back in time to Ancient Rome, and that they would understand, to teach them technologies which could prevent Rome’s collapse?

— oneusefulthing.substack.com

Treating housing as an investment was a colossal mistake

This is increasingly the issue of my generation in many countries and yet we don’t seem to be solving the supply/demand imbalance in housing almost anywhere.

Highlights

I hear advice from the older generation that “buying a house is an investment”.On one level, I get that. Buying a house is a better way to spend money than renting, because at the end, you have something. You have a whole frigging house.On another level, I think that the idea that home prices should outpace inflation is insane and maybe has broken modern society.

The current regime of “housing as an investment” rests on the idea that rising housing prices are good, because they generate wealth. And sure, a homeowner whose house value is rising fast enough is getting wealthier.But it’s a fake version of wealth generation. The majority of the higher value comes from demand outstripping supply. Little actual value is being created, only transferred. Those who own houses gain wealth at the expense of the people who don’t, either through higher prices when buying or renting.

For that reason — and I don’t wanted to come off as a polemicist here, but I gotta be honest — it’s not crazy to compare “housing as an investment” to a pyramid scheme. A pyramid scheme transfers money from the latecomers to the earlier adopters, often behind the facade of an allegedly amazing product. And while, according the proponents of the pyramid scheme, that product is the real wealth generator, very little wealth is being created. Instead, it’s a little bit of value creation and a whole lot of wealth transfer.

— goodreason.substack.com

Zap to heal wounds

This small, flexible, stretchable, wireless, battery-free device could help speed up the healing process.

Highlights

Not only can electricity drive electronics, it can help people heal. The first electronic bandage the body can absorb can apply electric signals to speed the healing of wounds by 30 percent, a new study finds.

— spectrum.ieee.org

It’s time to legalise sampling

It’s wild that you can cover a song but you cannot sample it. Hip hop has been around for fifty years now. It’s time to embrace this method of creating music.

Highlights

the album’s sound, sculpted largely by their producer Prince Paul, changed our expectations about what music itself could be: 24 riotous tracks—well-crafted songs interspersed with anarchic skits—made from hundreds of digital samples of sound from existing records. It was a “create first and ask permission later” kind of affair, as much of early hip-hop was. Because so many of those tiny digital samples had never been cleared with their original owners—either for digital distribution or for use of any kind—this foundational LP had never been available on apps like Spotify and Apple Music, rendering De La Soul’s oeuvre virtually invisible to later generations of listeners

The idea of legally recycling musical ideas has solid precedents. Songs are written, and once performed by a recording artist, remade by countless others. This process is completely legal, enshrined into U.S. law by what’s called a compulsory license: If you are the songwriter, you do not have the right to prevent anyone from rerecording your song, but if you are the recording artist, you must pay the songwriter an amount mandated by law

Yet there has never been a compulsory license for borrowing parts of songs, nor for pieces, or “samples,” of the recordings of those songs.

— slate.com

Could we make the web more immersive using a simple optical illusion?

A fun technical experiment from Shopify.

Meat-bag spaceships

This probably isn’t true, but I love this quote.

Highlights

the salinity of our bodies matches that of primordial seas, so in a sense we never really left the oceans.

Our micro-organic aquatic ancestors simply constructed meatbag spaceships with artificial life-support aquatic environments inside to explore beyond their oceanic home world.

— www.ribbonfarm.com

Anglish

As a lover of alternate history fiction, this is a fun project. What would English be like if it hadn’t borrowed so much from other languages?

Highlights

Anglish is a kind of English which prefers native words over those borrowed from foreign languages. Anglish is linguistic purism applied to English. For example: Dictionary > Wordbook Famous > Nameknown Native > Inborn Decide > Choose Computer > Reckoner This is achieved by simply choosing to use a native word over a borrowed word, or if there is no modern native word for a given concept, Old English words can be revived and updated to modern spelling and phonology to be used for a modern meaning

"We the Folk of the Foroned Riches, to make a more flawless oneship, build rightness, bring frith and stillness to our land, shield one another, uphold the overall welfare, and hold fast the Blessings of Freedom to ourselves and our offspring, do foresay and lay down this lawbook for the foroned riches of Americksland."

— anglish.org

Plastic roads

Plastic might be a cleaner, quieter, ready-made alternative to asphalt for the next generation of paving.

My biggest gripe with Australian urbanism is ugly roads and footpaths. These certainly look nicer.

Highlights

Today, bitumen is made from crude oil. It’s a sticky black product, the last bit out of the fractional process after petrol, diesel, butane, kerosene, and so on have all been extracted.

Proponents claim that plastic roads offer major improvements in design, cost, and quality while also serving as an outlet for waste plastic. Critics argue that plastic roads are empty ‘green’ hype or can’t compete with the benefits of traditional asphalt. The truth is most likely that using plastics in road building does have real environmental, cost, and structural benefits. But, outside of niches like bike paths and pedestrian paths, innovations still have yet to reach the scale and capabilities of the traditional asphalt road.

— worksinprogress.co

Sam Altman on planning for AGI and beyond

Most interesting takeaway for me is the insight that AGI is a section of the spectrum rather than a binary state.

Highlights

We want AGI to empower humanity to maximally flourish in the universe. We don’t expect the future to be an unqualified utopia, but we want to maximize the good and minimize the bad, and for AGI to be an amplifier of humanity. We want the benefits of, access to, and governance of AGI to be widely and fairly shared. We want to successfully navigate massive risks. In confronting these risks, we acknowledge that what seems right in theory often plays out more strangely than expected in practice. We believe we have to continuously learn and adapt by deploying less powerful versions of the technology in order to minimize “one shot to get it right” scenarios.

The “default setting” of our products will likely be quite constrained, but we plan to make it easy for users to change the behavior of the AI they’re using. We believe in empowering individuals to make their own decisions and the inherent power of diversity of ideas.

The first AGI will be just a point along the continuum of intelligence. We think it’s likely that progress will continue from there, possibly sustaining the rate of progress we’ve seen over the past decade for a long period of time. If this is true, the world could become extremely different from how it is today, and the risks could be extraordinary. A misaligned superintelligent AGI could cause grievous harm to the world; an autocratic regime with a decisive superintelligence lead could do that too.

Successfully transitioning to a world with superintelligence is perhaps the most important—and hopeful, and scary—project in human history. Success is far from guaranteed, and the stakes (boundless downside and boundless upside) will hopefully unite all of us.

— openai.com

Successful social networks start as a utility

Great quote from Kevin Systrom, founder of Instagram.

Highlights

I think all social networks start off as a utility and blossom into social networks long term, and that’s because they earn the right to be a social network

— stratechery.com

Doomers are wrong

Flamboyant pessimism serves no useful purpose.

Highlights

Even as capitalism has conquered the world, humanity is richer now than it has ever been, and for the last three decades income growth has been concentrated in low-income countries. Poverty is down at every level, not just in rates but in absolute numbers

Wage inequality is down a bit as well, with wages rising strongly for low earners since the mid-2010s. This is at least partly a result of the fact that practically everyone who wants a job in the U.S. has a job.

U.S. public social spending has risen steadily as a share of GDP. We have Social Security, SSDI, Medicare, Medicaid, unemployment insurance, Section 8 housing vouchers, SNAP, the EITC, the child tax credit, and many other safety net programs.

— noahpinion.substack.com

Cryptic protest icons

An interesting story about the history of avoiding censorship or persecution through the use of more cryptic icons for protest.

Highlights

In the 1980s a Polish anti-communist group called the Orange Alternative also used a seemingly random symbol to spread its message: a mythical creature with a tiny pointed hat. And that innocent image amplified a powerful political message to the world, which ultimately contributed to the fall of the Soviet Union.

Today, there’s a protest movement happening in Russia, which some people have compared to the Orange Alternative. It’s called The Little Picketers. The little picketers are small, clay figurines, about the size of the palm of your hand that are placed throughout Russian cities. Some of them hold peace signs, or Ukrainian flags, or anti-war messages. One of them is pictured holding a fish. It’s easy for anyone to get some clay and make a Little Picketer, and then discreetly drop it off in a public space without anyone else noticing. They usually get thrown away pretty quickly by Russian authorities — but before that happens, a photo is taken and submitted to an Instagram account, where they persist.

— 99percentinvisible.org

Build half a product, not a half-ass product

At one startup, we built an inventory management system as an additional product, ontop of our core functionality. We built support for purchase orders, stock adjustments, and stock taking.

I regret the way we built it. We should have built just one of those pillars (purchase orders for example) and gone really deep, rather than building a shallow version of all three.

By building all three before we launched the product, we launched three features that very few of our customers could use (because, thanks to the complexity of our core product, all of our customers had complex needs), when we could've launched one product that many of our customers would love, and then moved onto the other pillars later.

This meant that we were inundated with feature requests that we couldn't possibly juggle.

So, I agree with this principle from Basecamp. It's better to build one great feature than a few OK ones.

Highlights

Stick to what’s truly essential. Good ideas can be tabled. Take whatever you think your product should be and cut it in half. Pare features down until you’re left with only the most essential ones. Then do it again. With Basecamp, we started with just the messages section. We knew that was the heart of the app so we ignored milestones, to-do lists, and other items for the time being. That let us base future decisions on real world usage instead of hunches. Start off with a lean, smart app and let it gain traction. Then you can start to add to the solid foundation you’ve built.

What you really want to do is build half a product that kicks ass.

— basecamp.com

Accurately identifying users from only 100 seconds of motion data

These researchers were able to identify users with 94% accuracy from only 100 seconds of motion data from the Beat Sabre VR game.

One of the underrated consequences of AI is our future ability to fingerprint people with just about any data.

The cooling effect of irrigation

This study shows that large-scale irrigation can have a significant cooling effect. Perhaps part of our toolkit for adaptation as climate changes.

Highlights

We provide novel evidence that large-scale irrigation heterogeneously shifts the temperature distribution towards cooler temperatures during the months of the growing season relative to the rest of the year.

Cooling-by-irrigation propagates downwind and reduces the upper tail of the temperature distribution by up to 3C (5F) during the month of August, which has positive externalities on downwind crop yields ($120 million per year) and temperature-induced excess mortality ($240 million per year) that are of equal magnitude as the direct benefits of irrigation by enhancing heat tolerance ($440 million per year).

The observed cooling helps explain why the US has seen less warming, especially of very hot temperatures, than what climate models project. Our findings highlight that weather shocks in highly irrigated areas are not exogenous but are influenced by human responses in the form of irrigation.

— www.nber.org

The synthetic antibiotic that does not evoke bacterial resistance

A potentially crucial study that shows synthetic antibiotics could not only be effective but could solve the bacterial resistance problem.

Highlights

We identified a compound, COE2-2hexyl, that displayed broad-spectrum antibacterial activity. This compound cured mice infected with clinical bacterial isolates derived from patients with refractory bacteremia and did not evoke bacterial resistance.

These COE features enable the construction of a spectrum of compounds with the potential for development as a new versatile therapy for an imminent global health crisis.

— www.thelancet.com

Local-first software with CRDTs

This is a great explainer on how local-first applications can be architected. CRDTs are a technology I think most SaaS apps should consider at this point.

Highlights

CRDTs emerged from academic computer science research in 2011. They are general-purpose data structures, like hash maps and lists, but the special thing about them is that they are multi-user from the ground up.

If you are building a single-user application, you would maintain those data structures in memory using model objects, hash maps, lists, records/structs and the like. If you are building a collaborative multi-user application, you can swap out those data structures for CRDTs.

Users can view and modify the application state on their local device, even while offline. The CRDT keeps track of any changes that are made, and syncs the changes with other devices in the background when a network connection is available.

— www.inkandswitch.com

How does ChatGPT work?

A great article, as long as a short book, about how the technology behind ChatGPT works. Essential reading for anyone interested in this technology.

Highlights

That ChatGPT can automatically generate something that reads even superficially like human-written text is remarkable, and unexpected. But how does it do it? And why does it work?

The first thing to explain is that what ChatGPT is always fundamentally trying to do is to produce a “reasonable continuation” of whatever text it’s got so far, where by “reasonable” we mean “what one might expect someone to write after seeing what people have written on billions of webpages, etc.”

So let’s say we’ve got the text “The best thing about AI is its ability to”. Imagine scanning billions of pages of human-written text (say on the web and in digitized books) and finding all instances of this text—then seeing what word comes next what fraction of the time. ChatGPT effectively does something like this, except that (as I’ll explain) it doesn’t look at literal text; it looks for things that in a certain sense “match in meaning”.

— writings.stephenwolfram.com

Land tax skepticism

I support Georgist-style land taxes. Or, at least taxes on under utilised land. But, we really need to fix the planning red tape and the cost of construction first. Otherwise, a land tax could backfire as illustrated in this article.

Highlights

The levy, also called the land-value tax, is more radical than higher property taxes; it is an attempt to capture the entire value of land and redistribute it to the government and, in turn, the citizenry. As such, it requires separating the value of property improvements (such as buildings) and the value of the land itself.The theory is that land has nowhere else to go, so if you can tax the land value only, you can raise revenue without distorting the allocation of resources.

A land tax is only being talked about because urban planning is so broken, serving too many interests other than those of ordinary middle-class residents.

For a pure land tax to become reality, it would have to go through the meat grinder of local politics.I can predict what will come out of that meat grinder: a policy to compensate current landholders, one way or another, for the land tax. So if Palo Alto introduces a land tax, it is likely that the revenue will go back to those very same NIMBY interest groups.

— www.bloomberg.com

The personalities of Bing Chat

A great article from Ben Thompson on the personalities of Bing Chat.

One thing that has stuck with me:

Suppose these AIs are not conscious. When/if conscious AIs are eventually invented, what will be superficially different about them?

How will we recognise that an AI is conscious? This sounds challenging, given how conscious they already behave.

The story of Blake Lemoine, the Google engineer who publicly stated that LaMDA is sentient, seems a lot less ridiculous when I think about it this way. If it eventually does manifest, I suspect the people pointing it out will sound pretty crazy.

For the record, I believe they are not conscious, but I don't think we can authoritatively say so given we don't understand the mechanisms of consciousness.