Exponential View @exponentialview.

Joined July 2007
370 Photos and videos
Azeem Azhar retweeted
Extent to which White House allies are signaling that this is a culture war issue, not a technical one, is striking
I'm told Anthropic is perplexed by the situation they are facing, so they've turned to @k8em0 to do their on-the-record rapid response. These people really just don't get it....
19
79
552
203,410
Azeem Azhar retweeted
We just announced our Fusion API: - Fable-level performance on deep research tasks, at half the cost - Better-than-SOTA performance using panels The future of AI is neurodiversity, not single-model takeovers.
Introducing the Fusion API, the smartest compound model in the market. Fusion achieves Fable-level intelligence at half the price. How it works 👇
78
63
805
91,616
Azeem Azhar retweeted
Access to @AnthropicAI’s latest models has been paused for all customers, including in the US and UK. The main lesson: as we debate the future of national security and technological sovereignty, access to AI capabilities is crucial. That's why this Govt is the first to set up dedicated funding for our AI industry through @UKSovereignAI unit. And just this week, we announced £1.1bn for our AI chip industry.
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
154
103
974
295,783
Azeem Azhar retweeted
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees. The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance. Access to all other Claude models is not affected. We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible. Read our full statement: anthropic.com/news/fable-myt…
12,100
25,284
85,630
83,759,592
They won’t be free. because local work stations won’t copy with the volume of work we will send them.
Jun 12
Tokens will get 90% cheaper every year as models improve and $10k desktop workstations from @dell and @apple, running open source models, drive tokens to “essentially free” Token costs will be looked at like storage and bandwidth costs in a couple of years — which is to say you won’t think about them much.
2
1,257
Azeem Azhar retweeted
Realizing that Elon and I will have a combined net worth of $1.46 trillion this year
Jun 12
BREAKING: Our traders forecast Elon Musk to be worth $1.46 trillion this year — an all-time high
21
245
7,729
678,346
Seeing more on more pieces that flag as AI generated in mainstream publications. They should, at least, come with warnings. theguardian.com/commentisfre…
1
1
7
1,510
Wrote a book about this five years ago! amazon.com/Exponential-Age-A…
AI is advancing at a pace our policymaking institutions were never built for—and the gap between the two is becoming the central challenge of the technology. In his latest essay, our CEO Dario Amodei lays out how to close it. We're launching three new initiatives to support the efforts he outlines.
1
1
4
1,803
Azeem Azhar retweeted
On the tradeoff between AI vs. jobs: Several high-profile tech executives have said you should be spending as much on AI as you do on a software engineer’s salary…no one is actually doing that. The top 1% spends $7.45k per employee per month. The median firm is spending $11 per employee, about the cost of a single seat on a chat subscription. Not to say that won’t change – AI spend is still rising – but very few firms, if any, are spending that much. Despite their domination of the discourse.
18
11
79
8,240
Azeem Azhar retweeted
ai pilled companies spending $90k / employee per year 🤯 per employee, not per engineer. where does this go in a year!? can't wait to find out... from ramp's dataset (70k companies).
21
35
226
57,578
Azeem Azhar retweeted
NEW | Solar overtook coal generation in the US for the first month on record In May, solar generated an estimated 12.8% of US electricity, an all-time high! Coal produced 12.2% For now, it's one month, soon it will happen across the year.
14
92
330
34,604
Azeem Azhar retweeted
You can already start to see it.
Next is the part where solar pulls way ahead of the other two.
16
51
378
31,009
Azeem Azhar retweeted
UBS has a long report out: "Token Costs – What Customers are Saying" Companies are definitely feeling the higher bills, but the takeaway is that nobody’s actually slowing down AI use. They’re mostly just putting some guardrails in place and trimming spend in other areas while they keep rolling it out. In conclusion, UBS saying Google Cloud and AWS could end up with a cost edge here because of their own chips and models. That might help them take some share from Microsoft and the others over time. $GOOGL $AMZN $MSFT $ORCL
1
17
87
16,784
Azeem Azhar retweeted
$ORCL raised FY27 total revenue guidance to $90B, up from prior consensus of $86.6B, driven by surging demand for AI training and inferencing workloads. Management's tone remains exceptionally optimistic, describing the latest quarter as the strongest organic growth performance in over 15 years. However, they are also navigating complex execution challenges, including partner funding walkouts and data center delivery delays.
1
8
30
8,152
I mean, we gave companies legal personhood
We must not grant AI agents legal personhood ft.trib.al/BnEez25
1
1
4
1,396
Azeem Azhar retweeted
AI should earn its keep. Introducing the AI Productivity Guarantee. If Devin delivers less engineering value than you’re paying for, Cognition will fund your usage until it does, up to $10 million. It’s time for the AI industry to stop maximizing tokens and start maximizing productive output.
72
100
1,091
426,000
Napkin maths here The numbers seem to be hard to square up - last 12 months $90bn on gen AI ex China across consumers and businesses, so say 75% US and 75% corporate - about $45bn on gen AI by US corporates. Over the same period, the US wage bill was $11 trillion. US corporate revenues were $45 trillion. We're talking about 0.4% on GenAI as a portion of the labour bill or 0.1% of corporate revenues. Sure, the spend will be a power law with outlier firms spending OOMs more. But even Uber's seem closer to 1% of revenues. But plenty of studies (Bessen, 2020; Acemoglu, 2020) show that early automators build up persistent leads over later ones. So plausibly these firms have paid "over the odds" in return for learning and know-how that will compound over time.
For a couple of years, the main problem AI firms had was convincing people — or, more specifically, people in charge of companies — to pay for their products, writes tech columnist John Herrman. But following a period of rapid adoption, new issues have emerged. It makes sense “that companies with a general interest in AI as a labor-saving tool would underestimate just how costly going all-in on new AI tools actually is,” writes Herrman. “Compared to 2024, for example, the amount of compute used by the top-end models from Anthropic and OpenAI is astronomically higher in 2026. Concerns about cost are ‘all of a sudden a huge issue,’ Sam Altman says, but why wouldn’t they be? AI firms finally found a product corporate America wanted to buy. It just happened to be much more expensive than corporate America expected.” Read more from Herrman on the companies shopping around for their AI products: nymag.visitlink.me/jzXka3
2
2
19
10,756