LLMs are pretty good at giving you what you ask for. Not so good at telling you that you're asking for the wrong thing.

So they're comparable to rubber ducks. I would like to see data from a comparative study with rubber ducks, LLMs, and a control group.

Given a cheminformatics fingerprint definition based on SMARTS substructure patterns, come up with a screening filter, likely using a decision tree, which uses intermediate feature tests to prune search space faster than simply testing each pattern one-by-one.

For example, the Klekota-Roth patterns defined in their supplemental data (and also available from CDK at https://github.com/cdk/cdk/blob/main/descriptor/fingerprint/...) contain patterns like:

    "CC(=NNC=O)C",
    "CC(=NNC=O)C(=O)O",
    "CC(=NNC=O)C=C",
    "CC(=NNC=O)C=Cc1ccccc1",

Similarly, there are patterns like:

    "FC(F)(C=O)C1(F)OC(F)(F)C(F)(F)C1(F)F",
    "FC(F)(F)C(F)(F)C(F)(F)OC(F)(C=O)C(F)(F)F",
    "FC(F)(F)C(F)(F)C(F)(F)S",

So one stage might be to construct a list of element counts;

   ele_counts = [0]*200
   seen = set()
   for atom in mol.GetAtoms():
      ele_counts[eleno:=atom.GetAtomicNum()] += 1
      seen.add(eleno)

   ele_patterns = [
     # max known count, list of set of matching patterns
     (0, [set()]), # element 0
     (0, [set()]), # hydrogen
     ..
     (20, [{all patterns which contain no carbon},
           {all patterns which require at most 1 carbon}, ...
           {all patterns which require at most 19 carbons}],
     (10, [{all patterns which contain no fluorine}, ..
           {all patterns which contain at most 9 fluorines}], 
      ...]

   def get_possible_patterns(seen, ele_counts):
     for eleno in seen:
        max_count, match_list = ele_patterns[eleno]
        count = min(ele_counts[eleno], max_count)
        yield match_list[count]
   patterns = set.intersect(*get_possible_patterns(seen, ele_counts))

However, this is not sophisticated enough to identify which other tests, like the "CC(=NNC=O)C" example I gave before, or "S(=O)(=O)", which might be good tests at a higher level than the element.

And clearly if there isn't a sulphur, aren't two oxygens, and aren't two double bonds then there's no need to test "S(=O)(=O)", suggesting a tree structure would be useful.

The name was sticky enough that they've run with it, misunderstanding or ignoring that fundamental metaphor.

So if you’re streaming tokens from an LLM and you cancel the request from the client you’ll be wasting money.

I was one of the first to touch the OAI code model. Me and Albert developed the in-the-wild testing harness still in use today. We pulled all nighters to get GitHub approved for participation in the MSFT-OpenAI deal using those test results.

Existing Microsoft AI teams worked to halt our work, pushing their own small, worse models instead of OpenAI's.

I protoyped and lobbied for creation of the VScode extension. I invented and hacked the ghost text prototype into VScode, I invented the block based termination and implemented all the tree-sitter based logic needed to do it. Then I had to lobby up to Satya to get VScode to implement proper support in less than 6 months.

I named it Copilot.

I implemented GH auth, made the waitlist and onboarding. Helped design the e2e http/2 go server, after designing the fast.ly based precursor. Coordinated moving from OpenAI datacenter to Azure to improve Asia experience, and oversaw the cutover.

I was Chief Architect. It was my baby. Sad if this is how they are spinning the story internally at GitHub today.

Even though I don't use it personally, Copilot is a great tool and the bonus is laughable when you put it next to the impact it had on how people write code. For me lesson learned was that in order to spend your career in a good place and to save yourself months or years of time is to voice your concerns early and move on if all you hear are promises without any follow up.

That said, Alex's title appears to be principal engineer at GitHub, which is roughly equivalent to principal at Amazon in both expectations and comp. I worked at AWS for just shy of six years, and I can tell you that the person who created Firecracker -- which underpins huge portions of AWS's technology -- was at the same level, and while they were promoted shortly thereafter, they didn't receive a bonus (because comp packages at that level don't include bonuses at all). So, yeah, Alex is justified in sharing these details publicly, but this is just how it works in tech. (And, frankly, all the underpinning infrastructure that supported him wasn't his work -- he might have come up with Copilot, but would he have been able to without the work of thousands of others?)

He leveraged massive resources to do it and was allowed to tap people with multi-decades of experience (even implicitly - you think he build the execution engine where his "test harness" ran?). Not to mention that he essentially wrote a wrapper around an API..

If he put his total comp this pity party would shrink to one.

No matter the accomplishment, I doubt they would have been able to create copilot outside Microsoft and thus without Microsoft’s resources (computing infrastructure, engineering talent and software ecosystem).

And the “join a startup” is a dumb piece of advice as you’d be still making some one else profit.

Create your own company, and face the likelihood of bankruptcy.

I mean sure, technically this is correct. But it also fundamentally takes away from independent contributors and visionaries. Sure, "software is all about teams," but Linus Torvalds is definitely the guy behind Linux, and Palmer Luckey is the guy behind Oculus. It's unfair to take their achievements away because nowadays millions of people people contribute to Linux and Oculus was acquired by Meta.

I don't know OP, but I've seen this play out dozens of times in the cutthroat of corporate day-to-day, so it wouldn't surprise me for Microsoft to have a revisionist interpretation of how Copilot got started.

That's why the "I created" self-description rings hollow in my book -- as it does in essentially every other corporate / large-org case I've heard of (whether done as self-promotion, or the puff stories we're told about people we're supposed to hire or have run our teams or otherwise).

And which is why even people like Steve Jobs, who of course "was" the products he was famously associated with -- never went around saying "I created this." Not because he lacked for ego, or because his contributions were not monumental. But because he was smart enough to understand that once things get to a certain scale -- this kind of attribution is both absurd, and entirely beside the point.

I don't see this reading in the comment above. He is just sharing the passion and sense of ownership he put to create a great product. BTW He mentioned that it was a team effort in the thread, so I don't see any problem issue here.

John Carmack would agree with you.

And it's funny because while Carmack was that single person which was pretty much the best dev in the whole company, Romero was a better fit of your description.

I know I do :D

Microsoft may be bad but it's not the most evil company in history, not even by a long shot.

Exxon and Texaco (now Chevron) would also be high on the list.

I want to be clear: I am absolutely not giving Microsoft a pass here. They've done quite a lot of bad. But they're a relatively new kid on the block vis-a-vis other companies out there.

PS. Within tech, I would also say that Oracle might top the list of 'evil companies' - but in their case, they really seem to actively embrace it (not that that's a good thing).

I can't stand working with people like you, classic main character syndrome.

Edit: I’m not saying that MS / GitHub is obligated to do anything more than it already has. From a PR perspective, it’s just far cheaper to give a fatter bonus and promotion to early contributors to a big product and publicize it. It will also encourage other employees to push instead of just cruising. Otherwise, now you have someone making a logical case for not working very hard for either GitHub or MS

I am specifically referring to the features he lists in his comment above. E.g. ghost text, block completion, and OpenAI integration.

To add to his credit Copilot in VS proper still sucks, so I'd estimate original VS Code hack to have better skills than the whole (e.g. the sum of) VS proper Copilot extension team.

I made GitHub Copilot. Now I'm making an AI assistant on the web. Today it can order pizza, answer questions about articles, toggle settings, fill out forms, buy plane tickets, etc.

I need someone who understands the web deeply to help on frontend and internal tooling.

Join us to work at the bleeding edge of AI: prompting, fine-tuning, synthetic data, learning by example, codegen, planning, reasoning, and memory for embodied agents.

jobs+eng@minion.ai or DM @alexgraveley