Explore

Slice the corpus on any axis.

Filter by firm, year, function, metric or slide category — then see the result as a narrative arc, a treemap of slide types, or a catalog of slides.

clear
51,416 slides across 1229 decks match.
...and energy infrastructure begins to buckle
AirStreetCapital · industry_trends
...and energy infrastructure begins to buckle
5GW — Gigawatts
AI-first defense challengers scale, but are they exceptions?
AirStreetCapital · case_study
AI-first defense challengers scale, but are they exceptions?
$5.4B — valuation
AI shows promise on the frontline in Ukraine, but western hardware underwhelms
AirStreetCapital · case_study
AI shows promise on the frontline in Ukraine, but western hardware underwhelms
The debate over AI's economic impact intensifies
AirStreetCapital · context
The debate over AI's economic impact intensifies
Misinformation studies boom, but evidence of AI’s effectiveness remains thin
AirStreetCapital · key_messages
Misinformation studies boom, but evidence of AI’s effectiveness remains thin
Is AI going to be nationalized? (Spoiler alert: no)
AirStreetCapital · context
Is AI going to be nationalized? (Spoiler alert: no)
Safetyism to accelerationism: a major vibe shift has occurred
AirStreetCapital · context
Safetyism to accelerationism: a major vibe shift has occurred
OpenAI leadership struggle marks the start of an existential risk backlash
AirStreetCapital · case_study
OpenAI leadership struggle marks the start of an existential risk backlash
2023 Prediction: We see limited progress on global AI governance beyond high-level voluntary commitments
AirStreetCapital · key_takeaways
2023 Prediction: We see limited progress on global AI governance beyond high-level voluntary commitments
UK creates the world's first AI Safety Institute and the US swiftly follows
AirStreetCapital · context
UK creates the world's first AI Safety Institute and the US swiftly follows
Governments rush to patch gaps in critical national infrastructure
AirStreetCapital · industry_trends
Governments rush to patch gaps in critical national infrastructure
£59M — investment
Safety goes partisan (sort of)
AirStreetCapital · context
Safety goes partisan (sort of)
As the attack surface widens, developers up research into jailbreaking...
AirStreetCapital · industry_trends
As the attack surface widens, developers up research into jailbreaking...
...but they’re unable to keep up with the red team
AirStreetCapital · case_study
...but they’re unable to keep up with the red team
If you can’t beat the jailbreakers, join them
AirStreetCapital · case_study
If you can’t beat the jailbreakers, join them
It’s not just foundation models that face adversarial attacks
AirStreetCapital · case_study
It’s not just foundation models that face adversarial attacks
Beyond jailbreaking, research points to the potential of more stealthy attacks
AirStreetCapital · industry_trends
Beyond jailbreaking, research points to the potential of more stealthy attacks
Why is it so hard to predict the downstream capabilities of frontier models?
AirStreetCapital · diagnosis
Why is it so hard to predict the downstream capabilities of frontier models?
Although emergent capability scepticism is far from universal
AirStreetCapital · industry_trends
Although emergent capability scepticism is far from universal
Is RLHF breeding sycophancy?
AirStreetCapital · problem_statement
Is RLHF breeding sycophancy?
Direct Preference Optimization offers an escape from “reward hacking”…or does it?
AirStreetCapital · industry_trends
Direct Preference Optimization offers an escape from “reward hacking”…or does it?
RLHF isn’t going anywhere fast
AirStreetCapital · industry_trends
RLHF isn’t going anywhere fast
50-75% — Win rate against golden policy
Is a happy middle possible?
AirStreetCapital · case_study
Is a happy middle possible?
Can LLMs improve the reliability of...LLMs?
AirStreetCapital · diagnosis
Can LLMs improve the reliability of...LLMs?
Can LLM-generated critiques enhance both accuracy and alignment?
AirStreetCapital · industry_trends
Can LLM-generated critiques enhance both accuracy and alignment?
63% — preference rate
Can we make the known unknowns known?
AirStreetCapital · industry_trends
Can we make the known unknowns known?
70% — AUROC
Transparency is on the up, but there’s significant room for improvement still
AirStreetCapital · benchmark_peers
Transparency is on the up, but there’s significant room for improvement still
58 — Transparency Index Score
Could LLMs engage in 'reward tampering'?
AirStreetCapital · case_study
Could LLMs engage in 'reward tampering'?
Anthropic breaks open the black box...
AirStreetCapital · case_study
Anthropic breaks open the black box...
...and starts a trend for sparse autoencoders
AirStreetCapital · case_study
...and starts a trend for sparse autoencoders
7% — dead latents
Maybe the black box just isn't that opaque after all?
AirStreetCapital · case_study
Maybe the black box just isn't that opaque after all?
...but does this come with a downside?
AirStreetCapital · case_study
...but does this come with a downside?
Despite a storm of controversy, LLM biorisk remains uncertain...
AirStreetCapital · industry_trends
Despite a storm of controversy, LLM biorisk remains uncertain...
...but researchers point to other vulnerabilities
AirStreetCapital · problem_statement
...but researchers point to other vulnerabilities
Zooming out, are we too focused on the wrong harms?
AirStreetCapital · problem_statement
Zooming out, are we too focused on the wrong harms?
Follow our writing on AIR STREET PRESS (press.airstreet.com)
AirStreetCapital · closing_ask
Follow our writing on AIR STREET PRESS (press.airstreet.com)
A blueprint for addressing the global affordable housing challenge
McKinsey · 2015 · cover
A blueprint for addressing the global affordable housing challenge
By 2025, the affordable housing gap could affect 440 million households
McKinsey · 2015 · problem_statement
By 2025, the affordable housing gap could affect 440 million households
440 million — Households facing affordable housing challenge
We define the affordability gap as the difference between the annualized market price of a standard unit and the income available for housing
McKinsey · 2015 · problem_statement
We define the affordability gap as the difference between the annualized market price of a standard unit and the income available for housing
18 — Affordability gap
Global affordability gap is estimated to be about USD 650 billion
McKinsey · 2015 · market_sizing
Global affordability gap is estimated to be about USD 650 billion
USD 650 billion — Affordability gap
Building affordable housing could be an annual market of around USD 220 billion globally
McKinsey · 2015 · market_sizing
Building affordable housing could be an annual market of around USD 220 billion globally
$220 billion — Annual market size
A blueprint for tackling the affordable housing challenge
McKinsey · 2015 · framework_other
A blueprint for tackling the affordable housing challenge
Four levers can narrow the affordability gap
McKinsey · 2015 · section_divider
Four levers can narrow the affordability gap
Four levers can narrow the affordability gap
McKinsey · 2015 · framework_other
Four levers can narrow the affordability gap
100 — Annualized cost to own
$1 trillion public investment can unlock housing valued at $16 trillion
McKinsey · 2015 · impact_sizing
$1 trillion public investment can unlock housing valued at $16 trillion
$16 trillion — USD trillion
Land at the right location is the most critical lever
McKinsey · 2015 · section_divider
Land at the right location is the most critical lever
Hong Kong’s transit-oriented development unlocked land for new housing
McKinsey · 2015 · market_landscape
Hong Kong’s transit-oriented development unlocked land for new housing
600,000 — Affordable housing units
Urban greenfield development in outlying areas can fail if adequate infrastructure is not built
McKinsey · 2015 · problem_statement
Urban greenfield development in outlying areas can fail if adequate infrastructure is not built
14.0 — Ratio of vacant houses
Next page →