Phil @ContextSwitch

Last quarter I rolled out Microsoft Copilot to 4,000 employees. $30 per seat per month. $1.4 million annually. I called it "digital transformation." The board loved that phrase. They approved it in eleven minutes. No one asked what it would actually do. Including me. I told everyone it would "10x productivity." That's not a real number. But it sounds like one. HR asked how we'd measure the 10x. I said we'd "leverage analytics dashboards." They stopped asking. Three months later I checked the usage reports. 47 people had opened it. 12 had used it more than once. One of them was me. I used it to summarize an email I could have read in 30 seconds. It took 45 seconds. Plus the time it took to fix the hallucinations. But I called it a "pilot success." Success means the pilot didn't visibly fail. The CFO asked about ROI. I showed him a graph. The graph went up and to the right. It measured "AI enablement." I made that metric up. He nodded approvingly. We're "AI-enabled" now. I don't know what that means. But it's in our investor deck. A senior developer asked why we didn't use Claude or ChatGPT. I said we needed "enterprise-grade security." He asked what that meant. I said "compliance." He asked which compliance. I said "all of them." He looked skeptical. I scheduled him for a "career development conversation." He stopped asking questions. Microsoft sent a case study team. They wanted to feature us as a success story. I told them we "saved 40,000 hours." I calculated that number by multiplying employees by a number I made up. They didn't verify it. They never do. Now we're on Microsoft's website. "Global enterprise achieves 40,000 hours of productivity gains with Copilot." The CEO shared it on LinkedIn. He got 3,000 likes. He's never used Copilot. None of the executives have. We have an exemption. "Strategic focus requires minimal digital distraction." I wrote that policy. The licenses renew next month. I'm requesting an expansion. 5,000 more seats. We haven't used the first 4,000. But this time we'll "drive adoption." Adoption means mandatory training. Training means a 45-minute webinar no one watches. But completion will be tracked. Completion is a metric. Metrics go in dashboards. Dashboards go in board presentations. Board presentations get me promoted. I'll be SVP by Q3. I still don't know what Copilot does. But I know what it's for. It's for showing we're "investing in AI." Investment means spending. Spending means commitment. Commitment means we're serious about the future. The future is whatever I say it is. As long as the graph goes up and to the right.

What’s a startup conspiracy theory you believe? Mine: OKRs were actually a psyop from Google to slow down potential early stage competitors

@JimSproch @alexjlockwood @heylcs @Zhuinden @UrataSyndicate @Leeon4rdoo @codingisahabit @FcoPardoBaeza @VasiliyZukanov Sure it's possible, receivers are just objects after all. But there's a reason the context parameters KEEP exists - because receiver-heavy API design just aren't that ergonomic right now. Bundling a pile of pile of otherwise unrelated stuff into a single type == Context v2.

@JimSproch @alexjlockwood @heylcs @Zhuinden @UrataSyndicate @Leeon4rdoo @codingisahabit @FcoPardoBaeza @VasiliyZukanov It works fine for demos where you're replacing a single composition local, yes - lots of stuff looks good in trivial demos. What about the other 12? What if I want my composable to be an extension on `ColumnScope` or the myriad of other receivers I might want to use?

@JimSproch @alexjlockwood @heylcs @Zhuinden @UrataSyndicate @Leeon4rdoo @codingisahabit @FcoPardoBaeza @VasiliyZukanov But that doesn't really work until context parameters are a thing. All else being equal I'd like to get rid of the implicit dependencies too, but practically composition locals (used sparingly) don't cause the same problems regular globals do thanks to the "tree scoping".

One small positive among the disgusting scenes in Southport is this far right fuckwit getting his nadgers crushed by a flying brick.

If programmers invented mathematics, they'd have added operators called "super-plus" and "super-minus" which is addition and subtraction but with higher precedence than multiplication, "for people who prefer it that way", and then spent three decades trying to remove them.

@andrewhunterm I'm almost sure I heard @JamesHarkin laughing in the audience.

Apple Pie Position: A statement that instantly elevates the person who is saying it and is simultaneously hard for anyone else to push back on, and so everyone avoids the personal risk and just nods “yes”, even though its actual value in this specific situation might be relatively low, zero, or even negative. Because everyone in the workplace wants to come across as smart & competent, especially during meetings, Apple Pie Positions end up being frequently employed by your colleagues. Examples of Apple Pie Positions (trigger warning: might feel too painful at times) 1) “We need to define the success metrics for ” Also seen in the wild as: “How will you measure success here?” 2) “Do you have any data to back that up?” (narrator: it is either impossible to get useful data in this situation or the effort to get useful data would be greater than just doing it) 3) “We need more scalable team processes” 4) “We have too many meetings” 5) “We need to sync more often” 6) “This is a two-way door, so let’s just decide quickly” (narrator: it was actually a one-way door. or, it was a two-way door with a hungry lion on the other side of the door, so we kinda died) 7) “We need more metrics before making a decision here” 8) “We need a better go-to-market motion to improve product adoption” (narrator: the problem is you’ve built the wrong product, no amount of GTM sorcery is going to fix that) 9) “We need to clarify our expectations from people in this role so we can set them up for success” (narrator: the problem is that you’ve hired too many people whose work overlaps with other functions) 10) “We need better internal documentation!” (narrator: yes we do. we also need 50 other things so let’s just finally admit that this ain’t happening. besides OpenAI will solve this for us in 2024, right?) 11) “Let us not forget [quote one of the company’s core values].” (narrator: sometimes used as a weapon to justify one’s position, especially when faced with objective & rational counter-arguments for an issue) 12) “We need to replace our team meeting with written updates” Fast forward a few weeks/months: “People are missing live interactions with team members, they don't know what's going on across the team, nobody's reading the written updates. We need a meeting” 13) “Our current org structure doesn’t set us up for optimal execution of our strategic goals. That's why we are announcing a re-org” 14) “We must not slow down our shipping velocity” (narrator: usually applied to push back against a fairly rational measure that will cause a minor delay in shipping the product but will prevent bigger problems for the customers or the company further down the road) 15) “We need to post-mortem this” (narrator: you should have pre-mortem'ed this, so you wouldn't have had to deal with an ugly post-mortem. and the systemic changes rarely happen and somehow we are doing another ugly post-mortem 3 months from now. repeat) 16) “After 13 amazing months as CxO, Bob has decided to leave for other personal pursuits. The company wouldn’t be the same without his contributions” (narrator: this is not how it actually went down) 17) “We need an offsite so we can align the teams better with our strategic priorities” (narrator: we said the same thing for the previous 6 offsites. it almost seems like we are doing offsites more than we are being onsites) 18) “Remember, we are a mission-driven company” (narrator: particularly effective when justifying below-market salary / stock for current and prospective employees) 19) “We need to go on a customer listening tour to understand our customers better” (narrator: no, you need to pay more attention to the customer research that’s already happened over the past 18 months that we haven’t done anything about. you then need to translate that into differentiated features, instead of just building me-too tablestakes) 20) “We need more apple pie” (narrator: 🥧yum. yes we do)

@Piwai @_aurimas The public/private split is super painful for those outside Google. It's not uncommon for issues to be made private or closed as dupes of private issues. Links to private issues (or design docs) in CLs/commits/comments often make understanding the context behind changes hard too.

This is why you gotta promote on maintenance. Launching is easy, iterating and succeeding is hard (and where the money is made).

BuccoCapital Bloke @buccocapital

4 years ago

Reddit comment on why so many Google products get shut down “Internally, we called it the LPA cycle: Launch, Promo, Abandon”

92 1K 7K 0 885

@arriolachris FYI something's gone wrong with the code rendering in the article (there's no indentation or highlighting).

@PureGym Not at the new Hoxton Gym. There is only a Staff Wi-Fi at the moment.

@PureGym will there be Wi-Fi in the new Hoxton gym at some point? It's in a basement, so I don't get any 5G reception.

@zachklipp Not being able to match the line spacing behaviour of design tools. figma.com/blog/line-heig…

@and_kulikov @yigitboyar @intelligibabble @chrisbanes @askashdavies @objcode @chuckjaz Is there a reason source position can't work as a stable identifier for lazy lists, like it does for the rest of Compose? With the same caveats around explicit keys being required if a single source position can produce multiple items (e.g. a loop body).

@ianhlake Compose!

@Piwai My experience has been that the androidx + tools teams are pretty good at dealing with issues, but anything reported against the framework seems to largely go into a black hole.

@jdamcd The memory model really killed any enthusiasm I had for multiplatform - it might be good in theory, but it just doesn't work in an ecosystem where practically no existing code is written with its constraints in mind. The upcoming V2 is exciting though! blog.jetbrains.com/kotlin/2020/07…