It’s not gambling, these are legitimate financial instruments designed to allow proper risk management through appropriately market-set pricing on the value of that risk mitigation, and it’s doing this in a way that democratizes risk management in a way previously inaccessible to the public.
Employee enthusiasm isn’t really much of a factor. For better or worse, in the event of significant change of the brick & mortar day to day operations then employee continuity & institutional knowledge is even less of an actual strategic asset than the minimal treatment it already gets in consideration.
Depends on how market cap is defined for the purpose of the contract. Typical definition is just against floating shares in the market * share price. Debt doesn’t factor in at all except in so far as it will influence investor confidence -> share price.
That said: conceptually it’s not an awful fit for GameStop. In so far as video games discs and cartridges were the main disposable belonging i had as a kid and the main target for new purchases, Funcoland was (later to become GameStop), if you squint your eyes, a brick & mortar eBay scoped to only video games. If you’d been an SV startup at the time pitching the eBay concept you could have said “it’s like funcoland, but online and for anything and also lets people sell peer to peer “
LLM introspection tooling and observability, mainly around token-level behavior during inference; Also experimented with inference-time interventions (no retraining) e.g. improved DeBERTa on HANS via targeted layer/head adjustments.
I just put up a small demo of part of the tooling (HF Space):
-per-head projections into token space (logit lens-style)
-token rank changes across layers for target-tokens ("rank displacement")
-top-k next-token branches with internal state views
-mute a head at a given L x H coordinate
-inject tokens or rewind context
There's a minimal example in the UI showing how token candidates stabilize (or don’t) across layers
Background:
~15 years in data science/ analytics (higher ed), mostly translating technical work into decisions & policy for leadership.
More recently focused on LLM internals + tooling (Python/Rust, local model stacks, etc.) and agentic analytical tools for operational use.
Background in philosophy / applied linguistics & NLP; previously designed and taught a course on propaganda (how language shapes reactions to media etc.).
Looking for:
roles around LLM tooling, evals, interpretability, or applied AI where understanding model behavior is useful.
Tech: Python, Rust, SQL, embeddings, local LLM infra
And $ cost is a poor metric to chase when what you really care about includes a lot more-- exposure to the whims of geopolitical forces you can't foresee or control, which have both $ cost and more.
I agree to an extent… but a state forcing a nuclear share and locking the populace into higher power prices for 30+ years is going to politically very unpopular. Short term economic concerns dominate today.
Tough luck. That's the point of representative government: look out for the interests of the nation and sell it to your populace. If you can't sell it, be prepared to be voted out, but do the right thing.
Decommissioning always seemed odd absent either specific dangers or higher costs of operating than renewables. For new construction of course the costs shift dramatically, but existing plants that can continue to run would seem to provide exactly the legroom that enables more rapid expansion of renewables. Less time spent backfilling and exposure to both market and geopolitical forces of other energy sources, eg when there are disruptions of the sort going on now.
It isn't good optics at the moment, or good politics, for a company to loudly proclaim "we're firing people because of AI taking their jobs".
That doesn't mean that's what happened, it only means that whether or not its true, most companies aren't going to say it. The few that have said anything of the sort have suffered some backlash, and they aren't even as prominent as Meta or Microsoft (which also just announced plans to reduce by ~7% through buybacks, the first in their > 50 years) And this is on top of their decline to ~210,000 employees after 2025 firing of 15,000.
"When others explained that Facebook and Google didn't sell customer data, they didn't believe it"
I'm not sure there's a significant meaningful difference between direct selling and what they actually do, which is to make it available to target and manipulate people with extreme granularity. This is a huge part of why a person may not want their data to be held much less purchased to begin with, meaning it's "doesn't sell your data... but does or facilitates all of the things you do not want a group, in buying it from them, able to do."
It's a distinction without much practical difference.
Also: They buy your data from other brokers who do sell it, vastly enriching the degree to which customers of their ad platforms can make use of the data you already know they have far, far beyond your ability to know their full capabilities and the profile they have on you.
Again, it's not actually selling your data, but it's worth noting that when "they didn't believe it", that misconception was possibly helped along by Facebook or Google being on of the potential customers for that data either directly or via the proxy of a data broker whose largest customers are companies like that.
Selling your data means that anyone can have access your data forever. On the other hand, anyone can turn off ad personalization and delete their data on Google and Facebook.
A poor analysis that excludes any talk of the production costs of different aspects of getting a book to market.
Printing + labor in editing etc was 30% to 40% of sticker price in 1960. Today it's 10 to 15 or lower while author royties are about the same. Retail channels like Amazon demand a higher cut but not enough to cover that gap.
The net margin % of sticker paid by consumers above cost of production in all print and labor is paid by consumers, with the result that in raw dollars consumer still pay ~50% more than can be accounted for by inflation alone.
Consider: as a portion of average income, a very large # of everyday factory produced items are substantially less expensive, some even in non-inflation adjusted terms! A TV? A fraction of the cost. Plenty more examples. Books are expensive.
j/k totally gambling
reply