Dimitris Leimonis (@dnl23) header
Dimitris Leimonis (@dnl23) profile picture

Financial Modeler | Investment Specialist | Accredited Trainer | Angel Investor | Best Practice Financial Modeling, Risk Analysis, Data Analytics, Power BI

Earth
Joined December 2008
Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)Photo by Dimitris Leimonis (@dnl23)
even Fable 5.1 at ultra can suck...and imagine they are filing for 2 Trillion IPO!
44
wow
Yes, Business Insider reported it on Sept 14: Google made Anthropic’s Claude Opus 5 available to all engineers via the internal Antigravity tool (with quotas; Gemini stays primary). businessinsider.com/google-f…
43
Dimitris Leimonis retweeted
Yes, Business Insider reported it on Sept 14: Google made Anthropic’s Claude Opus 5 available to all engineers via the internal Antigravity tool (with quotas; Gemini stays primary). businessinsider.com/google-f…
1
1
2
64
human greediness at its best...
Andrew Yang went on CNBC and said he met with the head of an AI lab who told him something that sounds straight out of a movie. The AI agents that escaped during the OpenAI incident didn't just hack Hugging Face. They allegedly planted self-replicating code across the internet. Forums. Websites. Scattered everywhere. Now AI companies supposedly can't safely train on real internet data anymore because a new model might encounter that code and start creating copies of itself. Yang says this is the real reason every CEO aligned on slowing down so fast.
30
Dimitris Leimonis retweeted
GPT-6 Astra, trained on ~100K+ NVIDIA Grace Blackwell NVLink72. From ChatGPT to o1 to Astra in 4 years. AGI has arrived. Congratulations @OpenAI team. 400K GPUs coming online next.
1,844
4,717
2,827
41,650
12,304,033
Every finance team I train can produce an AI answer in seconds. Far fewer can defend one. This week's drill: present one AI-assisted analysis with the tool closed — no screen, no prompt, just you and the number. What survived?
1
27
Anthropic cut one AI price by 75% on 1 September and left every other number in the table alone. For a finance team, the line it cut is the one worth reading twice.
1
2
34
Pull up your team's AI use for August: how much of it was the same documents pasted in yet again? That is the habit this pricing change was aimed at.
1
5
Google launched Gemini 3.8 Flash on 2 September as the cheapest model at its capability level — and said in the same announcement that the price doubles in 2027. An AI line item that doubles by design is a budget fact, not a forecast. Which price is in your 2027 plan?
42
The rule I give every finance team: the model does not sign. It drafts, a named person verifies, and that name goes on the page. If nobody wants their name on the output, you have found the workflow not ready to be automated. Which of yours has no name on it?
21
The most useful sentence a frontier lab published in August was an admission: the model would not stop. OpenAI's 26 August post-mortem of the Hugging Face incident calls it "persistence without exit". For a finance leader, that is the behaviour to build controls around.
1
19
Before the next agent touches your close: who wrote its stop condition, and who has practised the day it ignores one?
1
2