Weeknotes 282 - ramping up for really helpful writing tools

This week's newsletter includes thoughts on the ongoing build-up to really helpful writing tools. In the news roundup, we have new AI battles, personal agents, LLM houses. And robots. A paper on the City as a Licence. And there are also some events to attend or track.

Weeknotes 282 - ramping up for really helpful writing tools
Midjourney - typing on a tablet with an AI ghostwriter as a partner represented as a digital cloud in the back of your mind

Hi all! Easter has just ended, but it was there when I was still in the middle of synthesising this newsletter. Next to that, I had to be careful about the latest news that was published on April 1: April’s Fool Day (Check how Gary Marcus is excited about the GPT-5 preview). This week's newsletter includes thoughts on the ongoing build-up to really helpful writing tools. In the news roundup, we have new AI battles, personal agents, LLM houses. And robots. A paper on the City as a Licence. And there are also some events to attend or track.

Triggered thought

Grammarly's new generative AI features are intriguing yet intrusive, raising questions about the balance between suggestion and takeover. Writing remains the signature case for generative AI, just as music was for recommender systems, with Spotify Discover Weekly as the prime example. In the realm of creative writing, the ideal tool acts as a buddy, writing coach, and background researcher, but it has yet to be perfected. Some tools, like Grammarly, focus on improving grammar and now aim to inspire better writing through generative AI. ChatGPT and Claude promise to help build stronger arguments by easily incorporating background sources. However, there is a tension: the writing produced by these tools, especially ChatGPT, can be cliché-ridden and uninspiring. Lex, a tool that has undergone iterations to enhance creativity in writing, may be worth revisiting, so I did.

While it doesn't yet add references, it is still necessary to construct narratives by prompting different services. I invited Lex to rewrite my first version, and it did a nice job, I think. I tried the rewrite function with Claude Opus and Sonnet, and with GPT-4. The latter creates rather formal speech and removes all personality. Funny enough, Claude Sonnet took a different standpoint, creating an observing piece: “the author wonders, etc…” So the above argumentation is built with the support of Claude Opus. And Grammarly, that keeps suggesting…

The question remains: when will we reach a point where a ghostwriter can start researching based on triggered concepts and collaboratively build a case without taking over the writing process entirely? One aspect is touched upon in the robotic facial expressions (see the news item below): in creating a natural feel of interaction, the AI must predict human behaviour to respond on time. But that is maybe something to dive into more at a later time and thought.

For the new subscribers or first-time readers, welcome! A short general intro: I am Iskander Smit, educated as an industrial design engineer, and have worked in digital technology all my life, with a particular interest in digital-physical interactions and a focus on human-tech intelligence co-performance. I like to (critically) explore the near future in the context of cities of things. And organising ThingsCon.

Notions from the news

Another round of news, of course, on AI, and more specific the generative AI modeling and tools. OpenAI had some new stuff again, to keep on track being leading the Generative AI course. Audio was this week's new iteration. Amazon is investing in Anthropic.

Navigating the Challenges and Opportunities of Synthetic Voices
We’re sharing lessons from a small scale preview of Voice Engine, a model for creating custom voices.

At the same time, GPT-4 is losing its position as the smartest kid in town. At least when it comes to being a creative writing buddy, Claude is taking over—for now.

GPT-4 loses its position as “best” LLM to Claude-3 in LMSYS benchmark
Grading large language models and the chatbots that use them is difficult. Other than counting instances of factual mistakes, grammatical errors, or processing speed, there are no…

A new kid in town is promising to be really open: DBRX by Databricks.

Introducing DBRX: A New State-of-the-Art Open LLM | Databricks

Summarising the last weeks of Apple AI rumours for WWDC. Wonder what is the update next week…

Apple isn’t getting a generative chatbot -- yet
WWDC 2024 is just a few months away, and the company is gearing up to discuss artificial intelligence — but one key Apple-made feature may not make its debut in iOS 18.

Let’s not forget the impact on a different level:

AI already uses as much energy as a small country. It’s only the beginning.
The energy needed to support data storage is expected to double by 2026. You can do something to stop it.

Trolling LLMs is the next step…

Grokking X.ai’s Grok—Real Advance or Just Real Troll?
Grok-1 is the largest open-source LLM yet, though not without caveats

And ads seem to become part of the business model. That might be problematic with product placements...

The ad model is coming to AI
Perplexity’s plan to add sponsored questions could be the future of chatbots — if it doesn’t destroy users’ trust

Adobe is going for a specific controlled strategy.

How Adobe’s bet on non-exploitative AI is paying off
The company says it’s proof that quality AI models don’t have to include controversial copyrighted content.

And politics is moving in too.

Inside the shadowy global battle to tame the world’s most dangerous technology
Can anyone control AI?

Will the GenAI bubble burst?

When Will the GenAI Bubble Burst?
Why and how it could happen in the next 12 months.

AI-gents

The other ‘battlefield’ is on the AI agents; last week, I shared the excellent post by Matt Webb; Every is diving into some backgrounds of the rising market.

What Are AI Agents—And Who Profits From Them?
The newest wave of AI research is changing everything

Chatbots are out of control, or are they taking more agency?

NYC AI Chatbot Touted by Adams Tells Businesses to Break the Law
The Microsoft-powered bot says bosses can take worker’s tips and that landlords can discriminate based on source of income. That’s not right.

On autonomous things

New York is getting Robotaxis, which are shaped like the shuttle concepts, rather than the usual self-driving cars.

New York City welcomes robotaxis — but only with safety drivers
NYC won’t permit fully driverless vehicles.

A house can be seen as a thing. Especially as it becomes an active player in your life using some LLMs. Or better, a office, a nice project by Harper Reed: the office as conversation partner. (via Monique)

Our Office Avatar pt 1: The office is talking shit again
I used sensors and an LLM to make my office talk. We used this to generate a humorous LLM-generated commentary - creating an interactive, profanity and personality-infused office space.

And robots

Round up on some robots and humanoids. The fastest.

This humanoid robot currently holds the world record for speed
The H1 is able to walk and run autonomously in complex terrains and environments.

This could be in the paper section, but let’s add it here: the human-robot facial coexpression quest. The lag in material behaviour shaping these facial expressions made the researchers create a system to predict human expressions based on the slightest facial changes.

Living together with humanoids. A conversation by two artists and a Sophia.

Misc

I agree with Patrick Tanguay (Sentiers) about LinkedIn's silent growing dominance as a social sharing place and the general insights from this article that extend beyond LinkedIn.

Paper for the week

An article discussing The City as a Licence as an alternative to The City as a Service, looking at blockchain and distributed ledger technologies. The city as a licence: Design, rights and civics in a blockchain society.

Through this lens, we frame Smart City platforms not as consumer service providers, but rather as “rights management systems” that provide or deny access to urban resources based on pre-set conditions and algorithmic decision-making.

(…)

The articles presented in this special theme serve as an initial step in this direction, but further investigation is urgently needed. While this special theme focused on the affordances of distributed ledgers, we are aware that many of the themes pertinent to Smart City urban governance also relate to other emerging technologies that allow for datafication, tokenization and algorithmically enabled rights management. Hence, we argue that The City as a License and the critical technical practices that emerge around it as a generative metaphor resonate with debates in the broader context of technologically mediated urban governance.

de Waal, M., Ferri, G., Gloerich, I., Vines, J., & Elsden, C. (2024). The city as a license: Design, rights and civics in a blockchain society. Big Data & Society11(1). https://doi.org/10.1177/20539517241227902

Have a great week!

The SHErobots-talks last week were very insightful, discussing our relations with robots. Nazli Cila was building a model on human collaborations described by Bratman, I hope to read more about. This shorter week is dedicated to a mix of program meetings on digital citizenship and commons. And some tech philosophy. Amsterdam UX goes good old behavioural design (BDAMS is still relevant apparently ;) ),

Next week, 9 April, IoT Day 2024, gathered some events around Europe; check the overview.

Enjoy your week!