Hi, I’m Azeem Azhar. This is a very special week for Exponential View: we’re celebrating the 100th edition of the Charts of the Week and the 400th edition of this Sunday Newsletter!

Before we get going, here’s a quick ask: readers of Exponential View make a big difference to our work when you 1) get a premium subscription — this funds our work, and 2) recommend the newsletter to your network. Join us in celebrating the 400th edition of EV by supporting our work in one of these two ways.

🧵 Share: Email | Linkedin | Twitter | Facebook | WhatsApp

Thanks!

In today’s edition:

The failure of voice assistants,
What to expect from GPT-4,
Longtermism is finally receiving the scrutiny it deserves.

Sunday chart: Silenced

As we cross the 400th edition of the Sunday newsletter today (hooray!), it’s a good time to look at one of the products many in the industry had high hopes for seven years ago when I started Exponential View. In 2016, Internet research legend, Mary Meeker, suggested that Amazon’s Echo smart speaker could be iPhone’s biggest threat. Indeed, on the surface, voice was looking like the next major thing, and Amazon Echo, with its smart assistant Alexa was leading the charge. In 2017, I mused that “Alexa is riding the wave of dramatic improvements in automated speech recognition driven by deep learning”. The number of skills — the things Alexa could do, often mostly provided by third parties like brands and media companies — grew 900% in two years. By some estimates about half of UK and US homes had a voice assistant of some sort last year.

Alexa was all the rage until it wasn’t. User activation and retention didn’t take hold. According to reports from Bloomberg last year, on many occasions “15% to 25% of new Alexa users were no longer active in their second week with the device”, and “most Alexa users in many years have used voice-powered devices only to play music, or set the timer while they cook, or turn on the lights.” The business of Amazon’s smart speaker was in activating users, monetising the services, and getting a kickback on the purchases made through the device. This didn’t happen.

The Alexa team is on track to lose Amazon $10bn this year. And the business is likely to be gutted.

I use voice systems: Siri, infrequently, for sending quick WhatsApp messages when I can’t get to the phone, and Otter daily for transcribing my thoughts.

But there is a myriad of concerns with voice assistants that really haven’t been dealt with in the past five years: having to remember a menu structure in my head; the lack of browsability; poor speech recognition; privacy concerns; and unsociability of barking at the machine over human conversation. In parallel the personalised interfaces we find on smartphones are getting better and better, making it harder for voice systems to stand out.

The voice assistant gold rush might be a template for other technologies. Voice is still on the wrong side of the uncanny valley but it was pushed by the large tech cos. These firms FOMO’d brands to build services for these assistants. Billions were spent and we ‘meh’d’. I’ve gotta say, the similarities with metaverse hype are inescapable.

On the generative wave, part 2: Waiting for GPT-4

It’s been 29 months since OpenAI launched GPT-2, its large language model which demonstrated the power of transformers-based neural networks. GPT-2 impressed with the quality of its natural text generation. Its successor GPT-3 — a bigger, more complex model — delivered even more powerful results.

Since GPT-3 was released in 2020, we have witnessed a wave of new innovations and products built on it and similar models. It isn’t only about feeding a model a text prompt and having it spew out lots of plausible-sounding (but not necessarily true) copy. We’re witnessing the same prompt-based approach to images, movies and more.

Now, rumours are swirling about GPT-4, OpenAI’s newest model due in the next three months. What should we expect?

I discuss this in my latest commentary, out tomorrow for paying members of Exponential View.

Subscribe now

Dept of our climate future

In every Sunday edition, we track key metrics that tell us a little about our shared climate future.

Our member, Marshall Kirkpatrick, takes the time to curate a view of our current climate status in this segment every week, and you can read Marshall’s view below:

One thing to know about our climate future this week: Philosophy matters and the stakes are high. After years of minimal critical engagement with the worldview known as Longtermism, advocated for by people like Sam Bankman-Fried and Elon Musk, a debate has finally begun. The Bulletin of Atomic Scientists tore Longtermism apart this week (via EV member Rafael Kaufmann). Put simply, longtermists seem to think climate change on Earth is a far smaller problem than whether humanity will transcend the Earth and colonise space. Instead of discounting the future, they discount the present, like an evangelical with eyes on a prize of eternal salvation.

Alternatively, this week I was introduced to the much more empowering strategy of the European Climate Knowledge and Innovation Community. Its mission “is to catalyse systemic change through innovation in areas of human activity that have a critical impact on greenhouse gas emissions and to create climate-resilient communities.” Unlike the space cowboys aiming to overshoot the present, this decade-old group builds “innovation ecosystems and community engagement activities; [to] source and exchange ideas; cultivate collaboration and attract new members to the community.” My favorite part is the group’s aim to “represent learning in the form of roadmaps for change.”

Now that’s an inclusive, responsible, and empowering narrative. And as Azeem has argued many times here at Exponential View: “Shifting the narrative around the clean transition is critical to driving more adoption and greater investment.”

In other climate developments:

Sponge cities are being built around China, led by one visionary landscape architect’s model of strategic water retention for climate adaptation and resilience.
Tidal power is dropping in price as turbines to harness this consistent source of renewable energy become more powerful and easier to deploy.
Renewable energy in the US is expected by year-end to represent 22% of national energy supply, exceeding coal (20%) and nuclear (19%). Wind and solar are up 18% year over year and 58% since 2019. Can the Inflation Reduction Act accelerate that growth enough to hit climate goals?
Big money: “Perhaps the biggest two months ever in the history of climate and energy finance” are what’s driving climate progress more than the frustrating COP processes, says Climate and Capital Media, citing $47 billion in decarbonisation deals announced in the past 30 days and $226 billion in the first half of 2022 (NEF).

Market data

More than two-fifths of American adults under 29 have bought a product after learning about it from an influencer.

Sales of electric scooters in France grew 42% in 2021

In China: 31% of all cars sold last month were battery EVs or plug-in hybrids

Average retail gas and electricity prices in Europe are up 144% and 79% on the long-term average.

Share

Short morsels to appear smart while naming extreme numbers

🗺️Precision positioning without satellites.

🙅India’s rage against the machine. via EV member Gianni Giacomelli

💡Extreme numbers are getting new names because we’re creating that much new data.

🧬The Great Depression left its mark in the DNA of people born during that recession.

📹A UK government ban on Chinese Hikvision CCTV cameras.

End note

I received a surprise this week: a copy of Volodymyr Zelenskiy’s new book, A Message from Ukraine. It is a compilation of his speeches since 2019. Zelenskiy is the most salient example of the power of oratory. It’s a quick and rather inspiring read, all the more poignant as Saturday 26th was Holodomor Memorial Day.

I also spent a few days in Venice at La Biennale. You can check out some of my photos of the incredible exhibits. I wanted to understand how contemporary artists are articulating today’s faultlines. I’ll be writing up some thoughts in the coming weeks, as they relate to climate change, the material cost of decarbonisation, perceptions of technology and the possibilities of science. It was also great to have dinner with some EV members in person: thanks to Paola, Marianna and Christina!

Ciao belli!
Azeem

7 years of Exponential View — our highlights:

Six ways coronavirus will change our world (Feb 5, 2020)
The negative cost of net zero (Sep 26, 2022) 🔒
The Russian hierarchy vs. the Ukrainian network (Mar 17, 2022) 🔒
A short history of knowledge technologies (Jul 21, 2020) 🔒
What complexity science tells us about job automation risk (Dec 18, 2019) 🔒

What you’re up to – notes from EV readers

Kevin Werbach explains why FTX’s failure is proof that crypto regulation works.

Simone Vannuccini and his co-author Maura Lombardi published “Understanding emerging patterns and dynamics through the lenses of the cyber-physical universe”, a paper conceptually bridging the ubiquity of information and computation, AI, decision-making, and tipping points in the global landscape.

David Coote wrote a report on application layer protocols used for energy divide demand management/flexibility, published by the International Energy Agency.

Jacob Huber is launching his group growth methodology combining empirically-supported theories with critical discussion and community.

Adam Oskwarek and his team at Zopeful are running a Black Friday campaign about the embodied emissions in everyday life, and providing a ready-made Carbon Removal Portfolio to take action.

To share your projects and updates, fill out your details here. Because of space constraints, we prioritise updates from paying members and startups I have invested in. (You can become the former by subscribing, if you have not already, and the latter by getting an intro to me via a trusted contact.)

Exponential View by Azeem Azhar

🔮Alexa flops; GPT-4; climate philosophy…