Forward Deployed Shitposter / Just midcurving / I own 0.1% of Personalis
Joined November 2022
- Tweets 3,531
- Following 1,700
- Followers 2,512
- Likes 9,103
833 Photos and videos
Pinned Tweet
27 Oct 2024
🫡
2
1
19
28,084
Great read on $GTLB Orbit. comparethemarketcareers.com/…
1
179
Great read on $GTLB Orbit. riftmap.dev/blog/gitlab-orbi…
2
183
You know you are on to something when not even all the engineers at Anthropic are fully aware yet that their own company has been cooking together with $GTLB. Cloning a git has become much faster & more efficient with @gitlab 3rd-gen-architecture.
Jun 15
It’s hard to imagine why cloning a git repo should be much slower than downloading an equivalent-sized file
Where are the experiments with custom git clients that clone faster
1
13
2,543
Jun 15
On my way to Frankfurt, testing @gitlab Orbit with SDLC context graphs & queries about vulnerabilities - works well on the train 🔥
Live demos in my session "What‘s new in GitLab: Live, unrehearsed and agentic" - tomorrow in Frankfurt, next week in Berlin. 🤗
Register below.
8
1,147
Jun 15
@AnjneyMidha, the guy who wrote the first check to Anthropic, shared an observation on feedback loops with @tracyalloway / @TheStalwart on Odd Lots recently:
"So when I say feedback, I'm in a very specific kind of feedback, which I call verifiable feedback. So when you say that wasn't right or that was wrong, that's an opinion, okay. Verifiable feedback is when you can have as close to factual verification as possible."
"So what does that actually look like?"
"In the case of software engineering, in the way software engineers actually code, is you write a piece of code and then you submit it to the main code base, and then you usually have a peer on your team review the code and approve it or reject it, and if it gets approved that's the first step that's called a PR a pull request. And if another human on your team that you trust approved it, that's one kind of verification of quality. And then two, before that piece of code usually gets deployed to a production system, you have unit tests, and those are quite objective tests of is this code performing the function we need it to, and if it passes both those tests, it's a verifiable piece of code that accomplished the goal. So in software engineering, the reason we've seen such a dramatic improvement and capabilities is that a lot of these labs are using feedback from that verification loop."
Sounds familiar? Anthropic is relying on $GTLB to host their code base. GitLab is providing a variety of tools for testing and quality controls that can provide verifiable feedback functions. @bstaples, is this a trend you see with customers?
Jun 14
The next big trade is infrastructure / RL environments that enable companies to turn their institutional knowledge / processes into continuously improving learning loops that they can own.
1
4
1,736
Jun 15
Sounds familiar to what @karpathy said few weeks ago too.
x.com/midwit_capital/status/…
Apr 29
@karpathy: "I thinks that there are some really valuable reinforcement learning environments that people could think of that are not part of the... yea I don't want to give away the answer...." (min. 14)
@bstaples, how about structured outcomes in $GTLB like pipeline pass/fail, MR approval/rejection, security scan positive negative?!
1
1
1,245
Jun 14
The next big trade is infrastructure / RL environments that enable companies to turn their institutional knowledge / processes into continuously improving learning loops that they can own.
28
32
590
170,625
Jun 12
Token-Efficiency-Maxxing, anyone?! The most interesting question is: to what extent can $GTLB capture the value of those saved tokens? If Claude Code with Orbit consumes 4X fewer tokens than Claude Code alone while also delivering better results, there’s really no rational reason not to use it.
Thinking this one step further: If the toll you charge on external agents to access your context graph is high enough (think value-based-pricing, not cost-plus) you can likely divert some usage from external agents to your own agentic offerings as they obviously wouldn't need to pay said toll?! Privileged position to be in.
Jun 12
The closest thing to magic you can find right now: lower agentic cost, improve outcomes, speed up reasoning - GitLab Orbit.
1
1
11
1,740
Jun 12
Incredible read on cancer vaccines.
Jun 8
How to build a cancer vaccine, and whether they will work this time
owlposting.com/p/how-to-buil…
for the past forty years, the medical field has attempted to create a 'cancer vaccine'. by and large, this effort has not been particularly successful. but things seem different this time around. there is, if you pay attention, a dizzying amount of optimism in the air regarding these potential miracle drugs. is it true? have we arrived? are cancer vaccines just over the horizon?
maybe! in this 8.3k word essay, i walk through the immunological theory behind cancer vaccines, the variants they come in, the (many) confusing aspects behind their few clinical successes, how machine-learning plays a role in their creation, and what lies ahead in their development
it's a very complicated subject, and i am deeply grateful to @iskander and @BenjaminGVincen for patiently answering my questions over the subject!
2
16
3,255
Jun 11
$GTLB big picture is shifting from DevSecOps to Agentic Infrastructure. Extremely comfortable holding my $1M position into this transition.
5
1
28
2,899
Jun 11
Couple of take aways from the $GTLB Transcend event.
Next-Gen Git: Re-wired architecture makes GitLab much faster & cheaper. In terms of the 'Lab Partnership', we didn't learn much besides the fact that Anthropic is impressed. It's in private beta. Details were very sparse. Two interesting comments from the Anthropic guy were that: (1) they just expect this stuff to work out of the box, and (2) for Fable-style large models this new architecture should be even more impactful.
2
1
26
1,726
Jun 11
$GTLB DAP: Recently made #5 on the @withmartian Code Review Bench. DUO-CLI recently crossed 100K weekly downloads. Customers saving time and seeing ROI using the product. Good stuff, still early.
1
2
607
Jun 11
New way of buying $GTLB via Flex. One commitment that can be spent across seats & tokens as demand changes. A plus is that eliminates the need for a new sales cycle once $GTLB launches new products (Orbit, Artefact Management, etc.) later this year. What I also think is true, is that Flex is somewhat specifically designed to seamlessly roll procured $GTLB seats into AI credits. Good response to the whole contracting seats debate. Everyone can move their contracts.
1
2
589