Fabijan Bajo

🛠️ Building | 12+ yrs in engineering & leadership

💡 Sharing experiences & learnings in tech

[Fabijan Bajo @FBajo]

There is a difference in code that passes tests an code that survives in production

[Fabijan Bajo @FBajo]

@Av1dlive I dont get it.. what makes the manual different for each user? can't you just share the manual?

[Fabijan Bajo @FBajo]

@Samaytwt man, i honestly thought this only happened on my mac due to indexing problems and my device always being full.

i just accepted it and assumed apple would never have such a shitty fuzzy search implementation..

i even thought they were hiding tools they didn't want me to open 🤣

[Fabijan Bajo @FBajo]

the filters-vs-collection-pages point is on point. a filter is just a url-params state a collection page is a real indexable node with its own copy and internal links. same underlying data, completely different SEO outcome.. you get the whole pyramid without authoring 40 pages by hand, and the head term ranks because the long tail under it does.did exactly this on a local directory..

[Developers @XDevelopers]

Announcing the hosted X MCP.

Agents now have access to the best real-time information source in the world.

Connect Grok, Cursor, or any MCP-compatible AI tool to the X API without any setup!

Check it out here: https://t.co/5MzPYwGFzD

[Anthropic @AnthropicAI]

Claude Fable 5 will be available again globally tomorrow.

After a series of productive conversations with the US government, we're redeploying the model with a new set of classifiers to target and block more cybersecurity tasks. In the near term, some routine tasks like coding and debugging will fall back to Opus 4.8. We’ll continue to refine these classifiers over the coming weeks to reduce false positives and better distinguish genuine misuse from legitimate requests.

We’ve also begun drafting a consensus framework—with Amazon, Microsoft, Google, and other Glasswing partners—for assessing the severity of AI jailbreaks and how AI developers should respond to them. We invite other industry partners and model providers to join us in this effort.

Finally, we’re scaling up our collaboration with the US government on model testing and safeguards. This will include pre-release access to models and safeguards for evaluation, information sharing on jailbreaks and misuse, and dedicated resources for joint research.

Thank you to our users for your patience, and to our partners across the government, industry, and the research community who worked alongside us to make Fable 5 available again.

Read our full blog: https://t.co/VHyum831ri

[Akhilesh Mishra @livingdevops]

Vibe coders everyday struggle in 15 seconds.

[Fabijan Bajo @FBajo]

@LeopoldATracker What is a good range

[Programmer Humor @PR0GRAMMERHUM0R]

hopeDrivenDevelopment

[Fabijan Bajo @FBajo]

https://t.co/Wg9FekLp0R

wow

[Fabijan Bajo @FBajo]

@claudeai @trq212 Lol all these complaints.. The hedonic treadmill is undefeated.

[Fabijan Bajo @FBajo]

@forgebitz buyer's lens: be wary of anyone telling you to dig when they sell shovels

maker's lens: be the one who digs. use your own tool til your hands blister, make it better

[Brian Armstrong @brian_armstrong]

Good take

My guess is
- demand for intelligence is near infinite
- but 80% of workloads will be running on 99% cheaper models within 12-18 months
- 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?)
- rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though
- this leads me to think the limiting factor will be energy and compute, not better models

At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially.

[Robert Greene @RobertGreene]

We are entering a world in which we can rely less and less upon the state, the corporation, or family or friends to help and protect us.

It is a globalized, harshly competitive environment.

We must learn to develop ourselves.

[Fabijan Bajo @FBajo]

Still can't get over the Chrome DevTools MCP. Letting an AI assistant simulate user behavior, capture screenshots, and live-debug console errors on localhost is an absolute game changer.

https://t.co/rae7FIBIKa

[Fabijan Bajo @FBajo]

Your CSV is a product. You just haven't shipped it yet.
ShipMore turns any spreadsheet into a live directory with search, SEO, and Stripe — built by an agent, not by hand.

https://t.co/O71YMHa9Bl

[Fabijan Bajo @FBajo]

Private evals are the new IP.

The real moat in the agentic era isn't the model you call—it's the proprietary evaluation suite, golden datasets, and guardrails that prove your agent can actually execute a task safely.

[Fabijan Bajo @FBajo]

honest question for anyone who's launched a directory: what killed it? my bet is data decay — every directory i loved went stale in 6 months. trying to solve exactly that.

[Fabijan Bajo @FBajo]

Reminder that asymptotic analysis isn't the whole story. Switching from Array-of-Structs to Struct-of-Arrays can net up to 30x on hot loops, and your total working set size determines whether random access hits L1 or DRAM. Excellent read:
https://t.co/FpevXX9OPE

[Aadit Sheth @aaditsh]

Andrej Karpathy shares a 3-step blueprint on how to master anything

[Fabijan Bajo @FBajo]

@martinfowler Agree; spawning agents isn't the skill. Designing the workflow around the one thing that can't scale "your own attention" is