Speed matters

Being 10x faster also changes the kinds of projects that are worth doing.

Last year I spent something like 100 hours writing a text editor. […] If I was 10x slower it would have been 20-50 weeks. Suddenly that doesn't seem like such a good deal any more - what a waste of a year!

It’s not just about speed of writing code:

When I think about speed I think about the whole process - researching, planning, designing, arguing, coding, testing, debugging, documenting etc.

Often when I try to convince someone to get faster at one of those steps, they'll argue that the others are more important so it's not worthwhile trying to be faster. Eg choosing the right idea is more important than coding the wrong idea really quickly.

But that's totally conditional on the speed of everything else! If you could code 10x as fast then you could try out 10 different ideas in the time it would previously have taken to try out 1 idea. Or you could just try out 1 idea, but have 90% of your previous coding time available as extra idea time.

Jamie’s model here helps explain the effect I described in AI-enhanced development makes me more ambitious with my projects. Prompting an LLM to write portions of my code for me gives me that 5-10x boost in the time I spend typing code into a computer, which has a big affect on my ambitions despite being only about 10% of the activities I perform relevant to building software.

I also increasingly lean on LLMs as assistants in the research phase - exploring library options, building experimental prototypes - and for activities like writing tests and even a little bit of documentation.

Via Reilly Wood

Tags: ai-assisted-programming, llms, ai, generative-ai

Like most academics these days, I spend a lot of time filling in online forms. Mostly, this is just an annoyance but occasionally I get something out of it. A recent survey in which the higher-ups tried to get an idea of how the workforce was feeling, asked the question “Do you think of the University as We or They?”.

Unsurprisingly given my reference to “higher-ups”, my answer was “They”. But giving the answer reminded me that, not so long ago, it would have been “We”. In its idealized form, a university was a self-governing community, with a well-understood teaching and research mission (which did not require a Mission Statement). All but the most senior management jobs were done by academics taking turns before returning to their real jobs. Administrative staff did essential work, largely independently, but didn’t conceive themselves as part of management.

The reality was inevitably less egalitarian and communitarian than this picture suggests, in all sorts of ways. Senior professors had too much power and inevitably, some of them abused it. And, given the times, lots of bad behaviour was tolerated that would not be now.

For good and ill, this has all been swept away, at least in Australia. Multiple layers of management are filled by people who have either left the academic life behind them or were never part of it. The university in this view, is not a community but a business enterprise, even if its ownership structure is rather opaque.

The reality is that of an ordinary workplace in which, most of the time, the interests of bosses and workers are in conflict (though, as in any workplace, there is a shared interest in the survival of the business). Senior managers see themselves as such and compare themselves to their corporate peers. Administrative job titles are those of the corporate sector (Chief Financial Officer and so on)>

Yet, as the question implies, there is a still a feeling that the university should be a We, and not merely in the sense of workers being willing to sing the company song. My own version of this is to think of the current regime as being temporary occupiers, from whom We will be liberated in due course. But others may take a more positive view – I’d be interested in commetnnts

It's a bit sad and confusing that LLMs ("Large Language Models") have little to do with language; It's just historical. They are highly general purpose technology for statistical modeling of token streams. A better name would be Autoregressive Transformers or something.

They don't care if the tokens happen to represent little text chunks. It could just as well be little image patches, audio chunks, action choices, molecules, or whatever. If you can reduce your problem to that of modeling token streams (for any arbitrary vocabulary of some set of discrete tokens), you can "throw an LLM at it".

— Andrej Karpathy

Tags: andrej-karpathy, llms, ai, generative-ai

There is superstition about creativity, and for that matter, about thinking in every sense, and it's part of the history of the field of artificial intelligence that every time somebody figured out how to make a computer do something - play good checkers, solve simple but relatively informal problems - there was a chorus of critics to say, but that's not thinking.

— Pamela McCorduck, in 1979

Tags: ai

One of the inclusion criteria for the 2024 Gartner Magic Quadrant for AI Code Assistants is... "at least 10 paying customer organizations."

Clearly everyone is using these things in the enterprise...

I'm just gonna say what a lot of us have been noticing this year.