ai-christianson

8 ms·

61.
by ai-christianson 1y ago
I've been building agents for a bit (RA.Aid OSS coding agent, now Gobii web browsing agents). The main problem with MCP is that it just makes tools available for the agent to use. We get the best performance when there's a small s
62.
by ai-christianson 1y ago
Websites should be able to request payment. Who cares if it is a human or an agent of a human if it is paying for the request?
63.
by ai-christianson 1y ago
It all comes down to the information content :)
64.
by ai-christianson 1y ago
We shipped RA.Aid, an agentic evolution of what aider started, back in late '24, well before CC shipped. Our main focuses were to be 1) CLI-first and 2) truly an open source community. We have 5 independent maintainers with full commit
65.
by ai-christianson 1y ago
I'm all for it. It enables those really big models with a relatively small number of active parameters.
66.
by ai-christianson 1y ago
IDK, modeling, constraints, simulations and stochastic processes seem like a match made in heaven. It's like how pairing a coding agent that can run unit tests and iterate is way more powerful than code gen alone.
67.
by ai-christianson 1y ago
The pre-red-tape internet was glorious. Only way to get that back is to decentralize everything.
68.
by ai-christianson 1y ago
> It’s amazing that this worked at all, but to be clear this layout is actually very bad. Just look at that minimum width trace used to carry power across the entire board and into the ESP32. Using min width traces and wrapping them and
69.
by ai-christianson 1y ago
Interesting, to me it's still very much a human in the loop process and the person whose name is on the commit is ultimately responsible for what they commit.
70.
by ai-christianson 1y ago
> The other side of it is people who seem to have 'gotten it' and can dispatch multiple agents to plan/execute/merge changes across a project and want to tell you how awesome their workflow is without actually showing
71.
by ai-christianson 1y ago
Right, I'm just pointing out that if you're "never going to get there anyway," then going 4x faster isn't going to help.
72.
by ai-christianson 1y ago
Moving very fast, but going nowhere.
73.
by ai-christianson 1y ago
We just shipped a major feature on our SaaS product. We, of course, used AI extensively. The thing is, this feature leaned on every bit of experience and wisdom we had as a team --things like making sure the model is right, making sure the
74.
by ai-christianson 1y ago
Classic network effects vs. product scenario.
75.
by ai-christianson 1y ago
> Yeah, it can look a bit repetitive if the code is already clear, but the context of why a thing is being done is still valuable. In the modern era with LLM tools, I'm sure it could be even more powerful. Is that because of literat
76.
by ai-christianson 1y ago
Why does this remind me of COBOL.
77.
by ai-christianson 1y ago
What's your favorite otel sink?
78.
by ai-christianson 1y ago
Is this better than Honeycomb?
79.
by ai-christianson 1y ago
He said "that cancelled" so I think it wasn't on his end.
80.
by ai-christianson 1y ago
All I know is I've been getting real world value out of LLMs since ~GPT3 and they've been producing more value with each release.
81.
by ai-christianson 1y ago
Might be a ridiculous question (I'm a software guy,) but is it at all possible to go the other way and increase the velocity of shipping and iterating on hardware to make it fit into the standard VC timelines?
82.
by ai-christianson 1y ago
But hardware is super capital intensive, right? So the right kind of investor could add significant value if they were aligned.
83.
by ai-christianson 1y ago
Anycast is pretty useful on the Internet :)
84.
by ai-christianson 1y ago
Employers are employees too
85.
by ai-christianson 1y ago
It's going to be wild with modern tech of 2038 but such an old school problem. I 100% think it will be a problem and will probably create a lot of jobs just like y2k fixing did... assuming jobs are still a thing.
86.
by ai-christianson 1y ago
Aren't the jobs they'll get be expecting them to use AI?
87.
by ai-christianson 1y ago
That seems way too good to be true. What's the catch?
88.
by ai-christianson 1y ago
Cool. Putting vision in the loop is a great idea. Ambitious idea, but I like it.
89.
by ai-christianson 1y ago
Is Zed managing the containerized dev environments, or creating multiple worktrees or anything like that? Or are they all sharing the same work tree?
90.
by ai-christianson 1y ago
> rather than looking at simulations You mean like automated test suites?

More ›