Fable 5.1 has a much better classifier for biology than when it came out.
It has no problem talking about the chemistry of cell signaling pathways.
When it first came out even basic biology questions were tripping the guard rails but not now.
Million? BILLIONS. So I ask myself: are they calling for a pause for the sake of mankind? Or are they proposing to put a throttle on their own gargantuan investment plan ?
Senior fullstack developer. Four years building enterprise ERP integration software for the energy, maritime and subsea sectors at a Norwegian B2B SaaS company. MSc in Computer Science, University of Stavanger.
Most of my work is the part other people avoid: integrating with accounting and ERP systems that have inconsistent APIs, undocumented edge cases and no usable test environment. Shipped production integrations against Tripletex, 24SevenOffice/Finago and Xledger. Built a drag-and-drop personnel planner on FullCalendar with shift and rotation logic, and label printing pipelines for Brady and Zebra hardware.
Before engineering I worked in product management, so I am comfortable scoping the problem with the customer and then building it. That tends to be useful on small teams where nobody wants to write a spec.
Looking for: contract or part-time, around 5 to 15 hours a week, alongside my current role. Overflow capacity for small consultancies and solo operators is exactly the kind of work I want. Open to full-time remote for the right role.
Comfortable with European hours, or US Eastern overlap in my evenings.
> anti-recording without policy for use and retention
Not recording on its own, but the linking of data with other cities and making it all searchable by those with access is where it becomes unprecedented mass surveillance.
You’re right that we have to be specific.
> your phone may be recording
They are able to dodge these accusations by just selling voice command logs. “No we don’t record you or sell your audio data” and yet Siri’s AUL reads:
[PhraseSpotter] Detected wake word “Pepsi” with a score 0.942.
[com.apple.voiceactivation] Proceeding with full wake.
A lot of people don’t know there is a low power “always on” processor (AOP) that monitors a small audio buffer on-device.
So no need to sell the audio, just the unrelated logs of some app that uses the AOP.
How do you do that in the current paradigm other than creating yet another gameable metric? And something I didn't mention above is that there is no difference between "solving the task" and "optimizing the metric" for an ML model, even though there clearly is for us. So it's not clear to me how you "fix" something that is baked into the architecture. All I'm saying is "instilling respect for human values" is not something that can actually be done via a cost function. In no small part because we humans probably don't even agree on those values, let alone on a single metric with which to quantify and "optimize" them.
For example, we agree that "merit" is valuable and that we should reward "merit." But to reward it we have to quantify it, and what metric should we use? Raw SAT score to get into college? But that also captures socioeconomic factors that unfairly penalize some and reward others. We generally agree that those who provide more value should earn more money, but what does that look like? Do we all agree on what activities are or should be valuable, or on how they should be rewarded? Until recently, I thought we all agreed that "empathy" was a human value, but a lot of people in this space, who are making these decisions unilaterally for all of us, don't apparently share that belief.
Suggesting every dark pattern for startups and then using record labels as the boogeyman like it's 2000 again is ... odd.
Why don't you let artists decide if they want record labels? They were far better off with EMI and Columbia records than with Spotify.
It is better to give $500,000 of $1,000,000 to a record label than $3 of $10 to Spotify. And record labels were pretty damn good at selecting artists, even in pop music. Does Paul Graham think music got any better after 2010?.
>when I was partially responsible for hardware refreshes in a UK public sector organisation with around a thousand endpoint PCs we budgeted something like £600 a box.
Same here in EU mainland, at every org I worked we used to get budget notebooks to do SW dev on, also in the 600-800 Euro range.
I think because the US is so much wealthier and can afford to spend more on tech without breaking a sweat. Hence the "Europoor" meme where Europeans can't comprehend the material abundance and spending power of the US.
Yep. Here’s an example of the output:
https://raw.githubusercontent.com/aeroscissorz/learnlance/re...
It’s easiest to understand by running it yourself, though. It’s available on PyPI:
pip install learnlance
Then:
learnlance setup
and start coding. The graph is generated from your coding session.
>Tempest Rising made more in profits than Marathon
Choosing the biggest flop as your benchmark is a very poor argument. If you lose half as much money as Marathon, you're technically twice as good, but it's still a flop and investors and publishers want successes, not half flops.
Marathon existing does not justify the investments into new RTS games. Quite the contrary. Marathon got hundreds of millions in cash because the shooter genre was making billions (at the time) so for them it was worth the gamble even if they lost. But the RTS genre is not making nearly enough money to justify Marathon levels of gambling on new IPs.
It started as license plate readers and expanded to include identifying people by clothing, body type, etc. storing it in a database that communicates with surveillance networks in other cities.
And how many copies are they gonna sell? How big is their player base?
All important things bean counters want to know before they pony up north of seven figures of investor money into another RTS and not into another looter shooter instead with a 100x revenue and 100x player base.
>Repairing a broken PC was a trivial help desk ticket; my PC was restored in an hour or two.
Except that even IT would know that a missing 8MB RAM stick is not a defect, it's (petit) theft, since RAM stick have no legs to just wander around by themselves. If it's gone, then an employee stole it.
Defect is when that 8MB stick would be ... defective.
That’s true, but the issue is still that we don’t have any LLM or any other kind of AI that follows laws perfectly to begin with. If we had AI that did exactly what we told it and nothing else I’d feel much more optimistic about things, I just think that will be a very hard task to achieve.
It's an echo chamber full of self-important piles of dogshit.
The real world is amazing... but this is a place for self-centered douche canoes like yourself to masturbate to your own imagined greatness. It's a place for assholes. It's a place for people whose fathers fucked up badly.
Claude responds with what things are not first. Even if reminded repeatedly.
Like Amodie, it serves to set the tone it "knows better" and then consumes the user's resources at an accelerated rate to try to correct it.
Fuck Anthropic, fuck Amodie, and fuck Claude. It's pretty obvious that consuming more tokens this way and making the user have higher cognitive load is a master class in extracting value from a system that is unsustainable.
Oh hey cool, didn't expect to see my project make Hacker News! I also develop little hardware devices to go with this stack if anyone is interested in embedded AppleTalk too. One is to connect your modern machine directly to LocalTalk networks, and the other is to emulate the now hard to find Apple network adapters for their 80s/90s printer line.
