aaronharnly

Well, this confirms that OpenAI have been shooting from the hip, not that we needed much confirmation. The fact that they repeatedly tried to hire Johansson, then went ahead and made a soundalike while explicitly describing that they were trying to make it be like her voice in the movie … is pretty bad for them.

I’m not sure if this is the one your instructor was referencing, but there was a fairly notorious such event in Afghanistan about a decade ago.

https://en.wikipedia.org/wiki/National_Airlines_Flight_102

https://www.faa.gov/lessons_learned/transport_airplane/accid...

As an aside, what is the current state of "LLMs for NPCs" in video games?

I don't really game or follow the industry, but I have to imagine both modders and publishers are working furiously to introduce more natural conversational experiences?

None of these are their core assets.

I like that "the heating of Earth's atmosphere by a hail of falling lunar debris would be destructive to all life" is only the second of three cited reasons for the proposal's rejection.

Natural disaster thingies

Thank you, great video.

I mean this feedback in the kindest way possible: you are taking a developer-out view of what a CRM is, and prioritizing the problems you perceived with Salesforce as a developer.

If you want to get a customer to pay money for your product, you really need to take a customer-in view. What problems do customers (some segment, perhaps a particular industry or country or type) have with buying Salesforce, and what can you do to make your product a better choice for them?

A flexible object data model and a UX that you as a developer find better are not likely to be those things.

According to this chart (whose source you have to pay to see, so unknown reliability), ATT Wireless has had well north of 150m customers in the date range in question, so it wouldn't be all of them.÷

And if they can’t say that (ie when it’s not true), they will intimate that you might be out of compliance, and hope you’re worried enough to just pay more anyway.

We use both Looker and Tableau. It’s astonishing to me how Looker has stagnated since being acquired by Google. It’s really moribund.

One does have to take into account that while once very respected, over the past decade Seymour Hersh has published a series of implausible and poorly sourced claims. That doesn’t mean this theory is necessarily false, but it doesn’t add a great deal of credence to it either.

My work password now has an "18" embedded somewhere in the middle of it thanks to my autoincrement approach to handling that kind of obnoxious policy.

Then I became CTO and retired the policy to align to modern NIST recommendations, so that "18" is in there forever :)

I visited a friend who was attending Deep Springs back in 1995 or 1996, when we were both first-year students.

Because I was too young to rent a car, I had to take a bus out of Las Vegas, and get picked up by someone from Deep Springs coming over the pass in a pickup. This is the Lida Junction mentioned in the article.

Very memorably, the "bus stop" was just an intersection with a dirt road, with (1) a phone booth (2) a brothel.

I have always complemented the three questions with a row-by-row look at the current stories or goals for the team, with the guiding question there being “what do we need to do to achieve this goal?”

The two are complementary in that the person-by-person polling helps flush out which people might need some extra help or swap around tasks.

Nice one! Gemini got it wrong in my (one) test:

https://g.co/gemini/share/4bae8ca3dd0a

Many wonderful paragraphs here, but to pull out perhaps a crux:

> If you were advertising a car, was it better to show a picture of a car, or of a pretty woman? If one was selling shirt collars, was it better to do so under the headline “Murder!” or the headline “Quality Shirt Collars”? To make a long story short (and to skip over the proto-pataphysical techniques by which Nixon established what we might call a “science of irrelevancy”), Nixon found that “irrelevant illustrations do attract more attention and hold interest longer than do relevant ones.” 8 Which is to say: systematically doing violence to sense raised revenue. This could be proven in his laboratory. Would it be wrong to say that we live, in many respects, in the world evolved from that finding?

The author is affiliated with the Friends of Attention, a very thoughtful and interesting collective of artists, academics, and others who are dedicated to helping reclaim our attention from its relentless “slicing, dicing, and pricing” by the Internet and media economy.

They have a terrific trove of artifacts and guides on their website:

https://friendsofattention.net/attention_trove

> ″The shift to a business unit operating model is a continuation of our client-focused, balanced execution priorities and is designed to accelerate product and services, go-to-market, and corporate functions’ operating efficiencies across all geographies we serve,” Xerox CEO Steven Bandrowczak said in the release.

Gosh, I really feel like I understand the rationale and the forward plan now.

The stated goal of Apollo was to land a man on the moon and return him safely to Earth.

Some of our customers (school districts) regularly recycle email addresses for students, reassigning an address to a new student after even just one year. That’s fine(ish?), but problematic when they don’t also provide a stable unique identifier that can signal the change. Identity management is a remarkably messy and hard problem — one of those areas where the “corner cases” add up to an order of magnitude more work than the core solution.

I don't have even a smidge of the background apparently required to make sense of this article. What is Netlify's battle with Vercel? What is their promotion and apparent now distancing from Jamstack about? Which frameworks would Netlify have ulterior motives to promote or disparage?

I haven’t, thanks! Signed up. Not wild about another layer of SaaS, but perhaps they’ll have a self hosted option.

Thank you, I appreciate the earnest and thoughtful reply!

Congratulations on shipping! We are currently evaluating replacing our homegrown version of an LLM proxy with this project:

https://github.com/BerriAI/litellm

Any comparison or contrast you would point out?

Based on some googling, it seems like Macromedia Director defaulted to 30 fps but could be customized for all or portions of a piece.

eg https://flylib.com/books/en/4.258.1.47/1/

I agree 75 would be surprising and I like your speculations!

Thank you!

Is there any solution to the recaptcha loop?

Huh! It has an image upload, and gives somewhat responsive, just not great, responses, so I'm a bit confused by that. So this is the existing Lens implementation?

I did some side-by-side comparisons of simple tasks (e.g. "Write a WCAG-compliant alternative text describing this image") with Bard vs GPT-4V.

Bard's output was significantly worse. I did my testing with some internal images so I can't share, but will try to compile some side-by-side from public images.