YouTube Hypes gives algorithmic powers to users
Plus: What AI work persona are you?
Hi product people 👋, Rich Holmes here. Welcome to the 370+ new subscribers who joined us since last week!
This week, I thought I’d try to re-jig the format slightly so I’ve replaced the longer introductory section with some more context and conversational elements throughout the other sections with a greater emphasis on new features and innovation. A few folks told us they’d prefer this type of format so I thought I’d try it out.
If you have any feedback on the re-format I’d be keen to hear your thoughts so just let us know.
Enjoy the rest of your week!
Reads and resources for product teams
Strategy - How to use generative AI for strategic planning
Harvard Business Review explores how some of the world’s top CEOs are using generative AI to power strategic planning - and highlight areas of potential growth that users may have missed. (Harvard Business Review)
UX and conversion - Surprising lessons from 20,000 product experiments
In this edition, Kyle Poyar explains what growth experiments not to run. Specifically, what you shouldn’t add to your website, statistically speaking. (Growth unhinged)
Fonts - A new fresh alternative to comic sans
If you’re a fan of the whimsical joy of comic sans but don’t feel entirely comfortable using it day today, this modern interpretation might be worth a look.
From the Department of Product Substack
Paid subscribers also get access to in-depth analysis of top companies through DoP deep dives and the Knowledge Series. Here’s some posts you may have missed from the Department of Product Substack:
How Uber drives engagement using a personalized notifications system
New product features and innovation this week
YouTube hosted its Made on YouTube event this week. At the event, it confirmed that it is testing a new feature that uses AI to generate ideas, titles and even full videos. Other announcements included plans to integrate Google’s most capable model for generating video, Veo, into YouTube Shorts that will allow users to generate backgrounds with generative AI, community hubs for superfans and a new interaction called ‘hypes’. Hypes allow users to give “hype points” to videos and the ones with the most hype points will increase their chances of discovery.
Jessica Locke, the product manager who worked on it, says that the reason they built the feature was because they discovered users wanted to help artists with discovery. It’s interesting to see the link between discovery algorithms and user interaction become so explicit in this way - definitely one to watch. You can read about the full product discovery process here.
Meanwhile…
OpenAI released its latest model, o1, which comes with advanced reasoning. It’s becoming increasingly difficult to keep up with some of these models’ capabilities so to help, here’s some practical, real-world examples of what it can do:
Write code in 1 hour that previously took 1 year - example
Code a first person shooter in HTML / JS - example
Creating a fully functional weather app - example
As more companies choose to adopt open source models instead, is this enough to maintain OpenAI’s dominance? Initial 2 -3 year enterprise contracts will be up for renewal soon so it’ll be interesting to see what renewal rates look like.
In other news…
DeepL has launched an AI glossary generator designed for business translations. Companies can now create custom translation glossaries by uploading files like PDFs or .docs. It automatically extracts relevant terms and their translations from uploaded documents, reducing manual work in glossary creation, which helps product teams to quickly expand internationally.
Retool has rebranded itself with a new makeover. The Retool product teams set out the design principles which underpinned the rebrand in a post entitled “The Art of Good Software”. Their design teams have done an excellent job at creating a modern, fresh new brand and kudos to them. PayPal have also unveiled a rebrand - but they seem to have lost their identity in the process.
Atlassian’s Jira is rolling out a new “global fields” feature which could be very helpful for product teams. It allows teams to create custom, global fields across multiple different assets like roadmaps or strategy docs.
Google is adding a new daily calendar overview in newly opened Chrome tabs for Workspace users with easy access to schedules and video calls. This is no doubt very helpful, but could potentially raise privacy concerns; if you’re sharing your screen on a work call and need to open a new tab, this could give other users a view of your upcoming appointments.
A new startup, Rep.ai, has raised $7.5 billion for its digital replica technology which helps tech support teams.
Tools you can use
AgendaHero lets you turn text, pdfs and other files into calendar events automatically.
Docs by Hashnode is designed for product teams to create Stripe-level quality API documentation for your product.
Hoop is a task manager which captures and prioritizes all of your tasks from Slack, Google Meet, Zoom, Gmail and more.
📈 Product data and trends to stay informed
A new Slack survey has categorized how people use AI at work for the first time.
The most common group is the Maximalists who are using AI multiple times a week and telling others about it. The Undergrounds use AI but are hesitant to tell their colleagues about it. The Rebels don’t subscribe to the AI hype and consider it unfair when coworkers use these tools. Superfans are excited about it but aren’t yet using it while Observers have yet to integrate AI into their work but are watching with caution and interest.
The Undergrounds strike me as an interesting group since they’re the ones using AI - but in secret. Which AI persona are you?
Netflix and YouTube are now the main rivals in the streaming war. Netflix posted revenues of $33.7 billion - just 7% higher than YouTube’s ad revenue last year (excluding YouTube’s Premium revenue).
Anthropic’s Claude now gets over 15 million web visits a month - impressive but still far behind ChatGPT’s 337 million.
Postings for software development jobs are down more than 30% since February 2020. The WSJ argues that engineering jobs have gone - and won’t be coming back.
A year long study has concluded that LLMs are now better at generating novel research ideas. In a product context, what impact could this have on the traditional role of user research?
Other product news in brief
LinkedIn has confirmed its AI models are trained on user data by default.
The entire board of 23andMe has resigned.
Amazon is mandating a 5 day a week return to office.
The Weekly Briefing is a product-led perspective of what’s happening in tech - and why it matters to product teams.
Paid subscribers also get access to in-depth DoP Deep dive reports to learn from the strategies and new features released by top tech companies, The Knowledge Series for sharpening your tech skills and more.