RIP Hyperloop, we hardly knew you.
Happy Friday! It’s the last Friday before Christmas. The Product Hunt team will be at a reduced capacity over the holidays but don’t worry. The digest will still hit your inbox.
Here’s some news:
🚆 One of the most promising Hyperloop startups is shutting down.
💬 Beeper may give up on its Android iMessage dream if the latest fix doesn’t work.
🍎 Apple has officially halted Apple Watch Series 9 and Ultra 2 online sales.
Midjourney just dropped a big update
Christmas came early for AI enthusiasts—the latest update to Midjourney, version 6, dropped yesterday. Built by the research team with the same name, v6 dropped as an alpha release with many improvements.
To start with, the way you prompt has changed according to the developers. Version 6 is much better at understanding you, so it’s important to be more explicit with your prompts. “V6 is MUCH more sensitive to your prompt. Avoid ‘junk’ like “award-winning, photorealistic, 4k, 8k” says Midjourney founder David Holz.
One of the biggest updates is how Midjourney handles text. You’ve probably seen it before; zoom in on an AI-generated image, and you’ll notice scrambled text. The team is tackling this head-on with what they call a “Describe” model, which helps the AI process text far better than previous iterations.
The team also updated how Midjourney processes things like lighting, shadows, and reflections. All of which are geared towards making images extremely photorealistic like the one above. Devs have had to up moderation systems and ask users not to “create images to cause drama.”
If you want to generate photorealistic images, just type "–style raw" at the end of your prompt.
Midjourney v6 is available for a short time. If you want to see what the hype is about, log in to Discord and enable it via the settings panel.
Your agents stopped answering and started doing
Your agents call tools, write to databases and move money, so a wrong answer is now a wrong action, which is a considerably worse meeting. Arcjet runs policy before every LLM call, tool call, query and API request: prompt injection, sensitive data, bots, rate limits, spend quotas, and your own rules in Rego on top. It reads the OpenTelemetry you already emit, so switching it on is one environment variable, no code changes and nothing to deploy. Each check also reads the earlier steps in the same workflow, so an agent's tenth action gets judged against the first nine and not on its own. JavaScript, Python and Go SDKs all at 1.0, agent framework guards included, free tier to start on.
Daily Top Products
