My annual MRI scan gives me a USB stick with the data, but you need this commercial windows software to open it.
Ran Claude on the stick and asked it to make me a html based viewer tool. This looks... way better.
one more prompt and it annotates everything with the findings
btw this is a good example of what i meant with "reflexively" reaching for AI. You tinker with AI for a while, and you just reach for this. This was an obvious thing to try when I saw I needed to use windows and was on my mac. You want to train your brain on this intuition.
Agents will become a common way people shop. So today we are releasing 3 tools to make adding commerce to those agents trivial:
- Checkout Kit: embed commerce widgets and checkout(!) directly into your agent and chat. This is already being used by Microsoft’s @Copilot.
- Shopify Catalog: low latency global product search across millions of merchants
- Universal Cart: shop from any store, anywhere, in one cart
With these, you can build seamless embedded commerce experiences quickly, starting with all the best products by the best merchants in the world instead of having to sign them all up individually. No need to build a complex new checkout, or dealing with regulatory marketplace rules. Commerce just plugs in and feels seamless in conversations.
We believe commerce should be everywhere people are, so we’re making that as easy as possible for the AI age.
There are some really cool technical details in how this all works. We heavily invested in MCP-UI, an extension of the MCP protocol to allow exposing widgets that a host can choose to display. The product cards you see in the video are exposed through MCP-UI and then securely rendered.
This means you don’t need to to fiddle with things like product variants, color swatches, subscriptions, minimums, bundles etc. The Web Components know about it all and render things correctly.
I just clued in how insane text2vid will get soon. As crazy as this sounds, we will be able to generate movies from just minor prompts and the path there is pretty clear.
Whisper allows very good transcription of existing videos and movies. Speaker detection is lacking but minor problem.
CLIP and Blip-2 are very good at extracting scene descriptions from still images, so you can also get set design, shot description and color grading.
When you feed scene transcript + scene description into GPT4 and prompt it to turn it into a movie script you get very good results. Also there are lots of movie scripts for real cinema floating around the internet which gets you ground truth.
Some really cool (programming) tech that I think are underrated right now and that have proven very useful to me this year:
Apache Arrow system, Parquet, Datafusion are incredibly strong foundation for the next chapter in data. DuckDB is a delight as well. We are going towards much simpler stacks based on extremely well designed infrastructure and that's really becoming for the data world.
Configuration files, json/yml/toml can get messy and the schema stories are complicated. Cuelang.org, while taking a moment to fully appreciate, is a fantastic tool to unify all those with automation build in and effortless value type based schema. Highly recommended.