Phil Tulip (@philtulip) profile picture

ai engineering

London, UK
Joined July 2026
Photo by Phil Tulip (@philtulip)
What happens when the cheap/subsidised tokens run out?
4
Something I'm encountering - developers can now deliver code quicker and earlier...but we still want code review and thorough QA before release. Testing smaller PRs but more of them...AI can easily chunk the work up but we're now stuck with more PRs to review. So, do we ditch code review? Not sure that can work in high regulatory environments. Interested to know views on this. The code review appears to handicap the ability to embrace AI (agentic) engineering.
9
Constantly in two minds about AI. On the one hand, it makes me more productive at my job (legacy processes not withstanding). On the other hand, anything I can now do with it, everyone else can. So the moat for any side project is non existent. And what I've observed both with myself and with others on the timeline - even if someone else has made a useful software/game/whatever, no one else really cares, and if they do, they try to build it themselves. As the tech gets better and this filters out to more of society, we'll see more and more people follow this customisation route. But as the exponential takes hold, I can only really see a small % of people bothering to continue to build and the rest will be mass consumers. Expect Ready Player One to become reality.
16
Opus 5.5 just makes it SO easy to create apps nowadays. I've just created two apps this past weekend that would have taken months previously...instead of feeling achievement, I just felt they were empty and disposable. Anyone else feel like this?
1
47
With artifacts, feels like all @AnthropicAI need to do is add custom domains/paths and a basic database and they have a lovable killer...do you even need version control? I have friends using artifacts directly as personal apps.
12
The great power of AI is teaching you things, I'm interested in (digital) filmmaking and wanted foundational knowledge of cinematography. So, one prompt later I have philtulip.com/learn/cinemato…
24
Resuming the Red Rising saga with Dark Age...it's been a few months reading other books (The Devils is great!) so I needed a refresher on the series and built this philtulip.com/red-rising/
16
We all know over consuming is bad - shorts, posts, TV, movies, music, etc But *reading fiction* is a positive consumptive activity. It forces (trains?) your brain to imagine sights and sounds. And imagination is the source of creativity. Read more fiction.
17
So much gold in this article and makes me think of the prominent AI engineers saying we should all be much more ambitious in our goals. A one person creative studio is certainly that!
16
Joining the vibecode game silliness... philtulip.com/strange-lights… Gotta catch em all! #threejs
27
Inspired by this approach to learning, built philtulip.com/how-the-world-… to summarise the fantastic book in an easy to follow, interactive digest. Buy the book, it's great!
We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks: Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better: Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better: Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better: Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work! In summary: - As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding. - Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.
19